I would just like to know the various AI algorithms or logics used in arcade/strategy games for finding/selecting best target to attack for individual unit.
Because, I had to write an small AI logic, where their will be group of unit were attacked by an various tankers, so i am stuck in getting the better logic or algorithm for selecting an best target for unit to attack onto the tankers.
Data available are:
Tanker position, range, hitpoints, damage.
please anybody know the best suitable algorithm/logic for solving this problem, respond early.
Thanks in advance,
Ramanand.
I'm going to express this in a perspective similar to RPG gamers:
What character would you bring down first in order to strike a crippling blow to the rest of your enemies? It would be common sense to bring down the healers of the party, as they can heal the rest of the team. Once the healers are gone, the team needs to use medicine - which is limited in supply - and once medicine is exhausted, the party is screwed.
Similar logic would apply to the tank program. In your AI, you need to figure out which tanks provide the most strength and support to the user's fleet, and eliminate them first. Don't focus on any other tanks unless they become critical in achieving their goal: Kill the strongest, most useful members of the group first.
So I'm going to break down what I feel is most likely pertains to the attributes of your tanks.
RANGE: Far range tanks can hit from a distance but have weak STRENGTH in their attacks.
TANKER POSITION: Closer tanks are faster tanks, but have less STRENGTH in their attacks. Also low HITPOINTS because they're meant for SPEED, and not for DAMAGE.
TANKER HP: Higher HP means a slower-moving tank, as they're stronger. But they won't be close to the front lines.
DAMAGE: Higher DAMAGE means a STRONGER tank with lots of HP, but SLOWER as well to move.
So if I were you, I'd focus first on the tanks that have the highest HP/strongest attacks, followed by the closest ones, and then worry about the ranged tanks - you can't do anything to them anyway until they move into your attack radius :P
And the algorithm would be pretty simple. if you have a list of tanks in a party, create a custom sort for them (using CompareTo) and sort the tanks by class with the highest possible HP to the top of the list, followed by tanks with their focus being speed, and then range.
And then go through each item in the list. If it is possible to attack Tank(0), attack. If not, go to Tank(1).
The goal is to attack only one opponent at a time and receive fire from at most one enemy at a time (though, preferably, none).
Ideally, you would attack the tanks by remaining behind cover and flanking them with surprise attacks. This allows you to destroy the tanks one at a time, while receiving no or little fire.
If you don't have cover, then you should use the enemy as cover. Move into a position that puts the enemy behind the enemy. This also improves your chance to hit.
You can also use range to reduce fire from multiple enemies. Retreat until you are only within range of one enemy.
If the enemies can all fire on you, you want to attack one target until it is no longer a threat, then move on to the next target. The goal is to reduce the amount of fire that you receive as quickly as possible.
If more than one enemy can fire on you at the same time, and you can choose your target, you should fire at the one that allows you to reduce the most amount of damage for the least cost. Simply divide the hit points by the damage, and attack the one with the smallest result. You should also figure in any other relevant stats. Range probably affects you and the enemy equally, but considering the ability to maneuver out of the way of fire, closer enemies are more harmful and should be given some weight in the calculation.
If moving decreases the likelihood of being hit, then you should keep moving, typically by circling your opponent to stay at their flank.
Team tactics would mostly include flanking and diversions.
What's the ammo situation, and is it possible to miss a stationary target?
Based on your comments it sounds like you already have some adhoc set of rules or heuristics to give you something around 70% success based on your own measures, and you want to optimize this further to get a higher win rate.
As a general solution method I would use a hill-climbing algorithm. Since I don't know the details of your current algorithm that is responsible for the 70% success rate, I can only describe in abstract terms how to adapt hill-climbing to optimize your algorithm.
The general principle of hill-climbing is as follows. Hopefully, a small change in some numeric parameter of your current algorithm would be responsible for a small (hopefully linear) change in the resulting success rate. If this is true then you would first parameterize your current set of rules -- meaning you must decide in your current algorithm which numeric parameters may be tweaked and optimized to achieve a higher success rate. Once you've decided what they are, the learning process is straight-forward. Start with your current algorithm. Generate a variety of new algorithms with slightly tweaked parameters than before, and run your simulations to evaluate the performance of this new set of algorithms. Pick the best one as your next starting point. Repeat this process until the algorithm can't get any better.
If your algorithm is a set of if-then rules (this includes rule-matching systems), and improving the performance involves reordering or restructuring those rules, then you may want to consider genetic algorithms, which is a little more complex. To apply genetic algorithms, it is essential that you define the mutation and crossover operators such that a single application of mutation or crossover results in a small change in the overall performance while a many applications of mutation and crossover results in a large change in the overall performance of your algorithm. I'm not an expert in this field but there should be much that comes up when you google for "genetic algorithms on decision trees". The pitfall to avoid is that if you simply consider swapping branches in a decision tree for the mutation operator, a single application might modify the root of your decision tree, generating a huge performance difference. This typically adds too much noise for a genetic algorithm, so my advice in this approach is to be very careful about the encoding of your operators.
Note that these two methods are very popular AI methods for learning or improving your current algorithm. You would do all of these simulations and learning offline. Then you would simply deploy the resulting, learned algorithm.
Related
I'm current using the expectiminimax algorithm, which is working great in my current situation:
max -> min -> chance -> max -> min -> chance -> (repeat)
I cannot in any way do
max/min -> Chance -> (repeat)
due to way the game works.
I feel as though the alpha is going to be inaccurate if I proceed to convert my algorithm over.
Is there any side effects to implementing pruning (apart from the horizon effect) with my current set up or am I just overthinking this?
I'm not sure which game you are trying to implement, but it does sound interesting. It's really hard to come up with a good answer when we know nothing about the branching factor of any of the nodes, or how the game proceeds. I think this is really depending on your game.
I've tried different pruning methods in the game of backgammon, and based on this experience I really think the result of your search algorithm depends on how much the chance node influence the game expectation compared to how much the min and max nodes changes the expectation.
If the chance node can change the expectation of the game drastically (high variance for each dice roll) but the five-six moves to choose among does not change the situation much, then I don't think you should worry to much.
However if it is the opposite, the chance nodes are basically just driving the game forward without influencing the game result expectation, and the game moves (or actions) really is important, I think you can gain a lot on getting a good search algorithm.
Check also *-minimax algorithm. (Bruce Ballard 1983)
The same things applies for Expectiminimax and *-minimax as normal minimax. Try the assumed best moves first (from some heuristics) to create cutoffs, but you should also try to heuristic order the chance outcomes in the chance nodes.
This is actually really interesting when you start testing this, but the only way to get good answers is to get you fingers dirty and try it out.
Good luck!
It appears that pruning still holds in this environment, I had serious doubts but everything has checked out so far. That is to say that all test cases in both the none pruning algorithm and the pruning algorithm provided the same move with the same heuristic just in fewer nodes.
I also found that in my case I would relinquish control of the chance nodes to the user (they can determine an acceptable loss if any for certain events). If the ability has a 95% chance to hit and a 5% chance to miss the user could deem that this 5% is negligible and ignore the miss chance, this does not mean the heuristic is treated as 100% of its value though, it maintains its 95%. I had rare situations where 16 possible outcomes could occur from using 1 ability given the environment and I wanted flexibility for such a situation.
As for the game, it would be similar to Pokémon in terms of play style at least.
I think I understand the basic concept of simulated annealing. It's basically adding random solutions to cover a better area of the search space at the beginning then slowly reducing the randomness as the algorithm continues running.
I'm a little confused on how I would implement this into my genetic algorithm.
Can anyone give me a simple explanation of what I need to do and clarify that my understand of how simulated annealing works is correct?
When constructing a new generation of individuals in a genetic algorithm, there are three random aspects to it:
Matching parent individuals to parent individuals, with preference according to their proportional fitness,
Choosing the crossover point, and,
Mutating the offspring.
There's not much you can do about the second one, since that's typically a uniform random distribution. You could conceivably try to add some random factor to the roulette wheel as you're selecting your parent individuals, and then slowly decrease that random function. But that goes against the spirit of the genetic algorithm and (more importantly) I don't think it will do much good. I think it would hurt, actually.
That leaves the third factor-- change the mutation rate from high mutation to low mutation as the generations go by.
It's really not any more complicated than that.
For my bachelor's thesis I want to write a genetic algorithm that learns to play the game of Stratego (if you don't know this game, it's probably safe to assume I said chess). I haven't ever before done actual AI projects, so it's an eye-opener to see how little I actually know of implementing things.
The thing I'm stuck with is coming up with a good representation for an actual strategy. I'm probably making some thinking error, but some problems I encounter:
I don't assume you would have a representation containing a lot of
transitions between board positions, since that would just be
bruteforcing it, right?
What could branches of a decision tree look
like? Any representation I come up with don't have interchangeable
branches... If I were to use a bit string, which is apparently also
common, what would the bits represent?
Do I assign scores to the distance between certain pieces? How would I represent that?
I think I ought to know these things after three+ years of study, so I feel pretty stupid - this must look likeI have no clue at all. Still, any help or tips on what to Google would be appreciated!
I think, you could define a decision model and then try to optimize the parameters of that model. You can create multi-stage decision models also. I once did something similar for solving a dynamic dial-a-ride problem (paper here) by modeling it as a two stage linear decision problem. To give you an example, you could:
For each of your figures decide which one is to move next. Each figure is characterized by certain features derived from its position on the board, e.g. ability to make a score, danger, protecting x other figures, and so on. Each of these features can be combined (e.g. in a linear model, through a neural network, through a symbolic expression tree, a decision tree, ...) and give you a rank on which figure to act next with.
Acting with the figure you selected. Again there are a certain number of actions that can be taken, each has certain features. Again you can combine and rank them and one action will have the highest priority. This is the one you choose to perform.
The features you extract can be very simple or insanely complex, it's up to what you think will work best vs what takes how long to compute.
To evaluate and improve the quality of your decision model you can then simulate these decisions in several games against opponents and train the parameters of the model that combines these features to rank the moves (e.g. using a GA). This way you tune the model to win as many games as possible against the specified opponents. You can test the generality of that model by playing against opponents it has not seen before.
As Mathew Hall just said, you can use GP for this (if your model is a complex rule), but this is just one kind of model. In my case a linear combination of the weights did very well.
Btw, if you're interested we've also got a software on heuristic optimization which provides you with GA, GP and that stuff. It's called HeuristicLab. It's GPL and open source, but comes with a GUI (Windows). We've some Howto on how to evaluate the fitness function in an external program (data exchange using protocol buffers), so you can work on your simulation and your decision model and let the algorithms present in HeuristicLab optimize your parameters.
Vincent,
First, don't feel stupid. You've been (I infer) studying basic computer science for three years; now you're applying those basic techniques to something pretty specialized-- a particular application (Stratego) in a narrow field (artificial intelligence.)
Second, make sure your advisor fully understands the rules of Stratego. Stratego is played on a larger board, with more pieces (and more types of pieces) than chess. This gives it a vastly larger space of legal positions, and a vastly larger space of legal moves. It is also a game of hidden information, increasing the difficulty yet again. Your advisor may want to limit the scope of the project, e.g., concentrate on a variant with full observation. I don't know why you think this is simpler, except that the moves of the pieces are a little simpler.
Third, I think the right thing to do at first is to take a look at how games in general are handled in the field of AI. Russell and Norvig, chapters 3 (for general background) and 5 (for two player games) are pretty accessible and well-written. You'll see two basic ideas: One, that you're basically performing a huge search in a tree looking for a win, and two, that for any non-trivial game, the trees are too large, so you search to a certain depth and then cop out with a "board evaluation function" and look for one of those. I think your third bullet point is in this vein.
The board evaluation function is the magic, and probably a good candidate for using either a genetic algorithm, or a genetic program, either of which might be used in conjunction with a neural network. The basic idea is that you are trying to design (or evolve, actually) a function that takes as input a board position, and outputs a single number. Large numbers correspond to strong positions, and small numbers to weak positions. There is a famous paper by Chellapilla and Fogel showing how to do this for a game of Checkers:
http://library.natural-selection.com/Library/1999/Evolving_NN_Checkers.pdf
I think that's a great paper, tying three great strands of AI together: Adversarial search, genetic algorithms, and neural networks. It should give you some inspiration about how to represent your board, how to think about board evaluations, etc.
Be warned, though, that what you're trying to do is substantially more complex than Chellapilla and Fogel's work. That's okay-- it's 13 years later, after all, and you'll be at this for a while. You're still going to have a problem representing the board, because the AI player has imperfect knowledge of its opponent's state; initially, nothing is known but positions, but eventually as pieces are eliminated in conflict, one can start using First Order Logic or related techniques to start narrowing down individual pieces, and possibly even probabilistic methods to infer information about the whole set. (Some of these may be beyond the scope of an undergrad project.)
The fact you are having problems coming up with a representation for an actual strategy is not that surprising. In fact I would argue that it is the most challenging part of what you are attempting. Unfortunately, I haven't heard of Stratego so being a bit lazy I am going to assume you said chess.
The trouble is that a chess strategy is rather a complex thing. You suggest in your answer containing lots of transitions between board positions in the GA, but a chess board has more possible positions than the number of atoms in the universe this is clearly not going to work very well. What you will likely need to do is encode in the GA a series of weights/parameters that are attached to something that takes in the board position and fires out a move, I believe this is what you are hinting at in your second suggestion.
Probably the simplest suggestion would be to use some sort of generic function approximation like a neural network; Perceptrons or Radial Basis Functions are two possibilities. You can encode weights for the various nodes into the GA, although there are other fairly sound ways to train a neural network, see Backpropagation. You could perhaps encode the network structure instead/as well, this also has the advantage that I am pretty sure a fair amount of research has been done into developing neural networks with a genetic algorithm so you wouldn't be starting completely from scratch.
You still need to come up with how you are going to present the board to the neural network and interpret the result from it. Especially, with chess you would have to take note that a lot of moves will be illegal. It would be very beneficial if you could encode the board and interpret the result such that only legal moves are presented. I would suggest implementing the mechanics of the system and then playing around with different board representations to see what gives good results. A few ideas top of the head ideas to get you started could be, although I am not really convinced any of them are especially great ways to do this:
A bit string with all 64 squares one after another with a number presenting what is present in each square. Most obvious, but probably a rather bad representation as a lot of work will be required to filter out illegal moves.
A bit string with all 64 squares one after another with a number presenting what can move to each square. This has the advantage of embodying the covering concept of chess where you what to gain as much coverage of the board with your pieces as possible, but still has problems with illegal moves and dealing with friendly/enemy pieces.
A bit string with all 32 pieces one after another with a number presenting the location of that piece in each square.
In general though I would suggest that chess is rather a complex game to start with, I think it will be rather hard to get something playing to standard which is noticeably better than random. I don't know if Stratego is any simpler, but I would strongly suggest you opt for a fairly simple game. This will let you focus on getting the mechanics of the implementation correct and the representation of the game state.
Anyway hope that is of some help to you.
EDIT: As a quick addition it is worth looking into how standard chess AI's work, I believe most use some sort of Minimax system.
When you say "tactic", do you mean you want the GA to give you a general algorithm to play the game (i.e. evolve an AI) or do you want the game to use a GA to search the space of possible moves to generate a move at each turn?
If you want to do the former, then look into using Genetic programming (GP). You could try to use it to produce the best AI you can for a fixed tree size. JGAP already comes with support for GP as well. See the JGAP Robocode example for an instance of this. This approach does mean you need a domain specific language for a Stratego AI, so you'll need to think carefully how you expose the board and pieces to it.
Using GP means your fitness function can just be how well the AI does at a fixed number of pre-programmed games, but that requires a good AI player to start with (or a very patient human).
#DonAndre's answer is absolutely correct for movement. In general, problems involving state-based decisions are hard to model with GAs, requiring some form of GP (either explicit or, as #DonAndre suggested, trees that are essentially declarative programs).
A general Stratego player seems to me quite challenging, but if you have a reasonable Stratego playing program, "Setting up your Stratego board" would be an excellent GA problem. The initial positions of your pieces would be the phenotype and the outcome of the external Stratego-playing code would be the fitness. It is intuitively likely that random setups would be disadvantaged versus setups that have a few "good ideas" and that small "good ideas" could be combined into fitter-and-fitter setups.
...
On the general problem of what a decision tree, even trying to come up with a simple example, I kept finding it hard to come up with a small enough example, but maybe in the case where you are evaluation whether to attack a same-ranked piece (which, IIRC destroys both you and the other piece?):
double locationNeed = aVeryComplexDecisionTree();
if(thatRank == thisRank){
double sacrificeWillingness = SACRIFICE_GENETIC_BASE; //Assume range 0.0 - 1.0
double sacrificeNeed = anotherComplexTree(); //0.0 - 1.0
double sacrificeInContext = sacrificeNeed * SACRIFICE_NEED_GENETIC_DISCOUNT; //0.0 - 1.0
if(sacrificeInContext > sacrificeNeed){
...OK, this piece is "willing" to sacrifice itself
One way or the other, the basic idea is that you'd still have a lot of coding of Stratego-play, you'd just be seeking places where you could insert parameters that would change the outcome. Here I had the idea of a "base" disposition to sacrifice itself (presumably higher in common pieces) and a "discount" genetically-determined parameter that would weight whether the piece would "accept or reject" the need for a sacrifice.
I'm wondering how people test artificial intelligence algorithms in an automated fashion.
One example would be for the Turing Test - say there were a number of submissions for a contest. Is there any conceivable way to score candidates in an automated fashion - other than just having humans test them out.
I've also seen some data sets (obscured images of numbers/letters, groups of photos, etc) that can be fed in and learned over time. What good resources are out there for this.
One challenge I see: you don't want an algorithm that tailors itself to the test data over time, since you are trying to see how well it does in the general case. Are there any techniques to ensure it doesn't do this? Such as giving it a random test each time, or averaging its results over a bunch of random tests.
Basically, given a bunch of algorithms, I want some automated process to feed it data and see how well it "learned" it or can predict new stuff it hasn't seen yet.
This is a complex topic - good AI algorithms are generally the ones which can generalize well to "unseen" data. The simplest method is to have two datasets: a training set and an evaluation set used for measuring the performances. But generally, you want to "tune" your algorithm so you may want 3 datasets, one for learning, one for tuning, and one for evaluation. What defines tuning depends on your algorithm, but a typical example is a model where you have a few hyper-parameters (for example parameters in your Bayesian prior under the Bayesian view of learning) that you would like to tune on a separate dataset. The learning procedure would already have set a value for it (or maybe you hardcoded their value), but having enough data may help so that you can tune them separately.
As for making those separate datasets, there are many ways to do so, for example by dividing the data you have available into subsets used for different purposes. There is a tradeoff to be made because you want as much data as possible for training, but you want enough data for evaluation too (assuming you are in the design phase of your new algorithm/product).
A standard method to do so in a systematic way from a known dataset is cross validation.
Generally when it comes to this sort of thing you have two datasets - one large "training set" which you use to build and tune the algorithm, and a separate smaller "probe set" that you use to evaluate its performance.
#Anon has the right of things - training and what I'll call validation sets. That noted, the bits and pieces I see about developments in this field point at two things:
Bayesian Classifiers: there's something like this probably filtering your email. In short you train the algorithm to make a probabilistic decision if a particular item is part of a group or not (e.g. spam and ham).
Multiple Classifiers: this is the approach that the winning group involved in the Netflix challenge took, whereby it's not about optimizing one particular algorithm (e.g. Bayesian, Genetic Programming, Neural Networks, etc..) by combining several to get a better result.
As for data sets Weka has several available. I haven't explored other libraries for data sets, but mloss.org appears to be a good resource. Finally data.gov offers a lot of sets that provide some interesting opportunities.
Training data sets and test sets are very common for K-means and other clustering algorithms, but to have something that's artificially intelligent without supervised learning (which means having a training set) you are building a "brain" so-to-speak based on:
In chess: all possible future states possible from the current gameState.
In most AI-learning (reinforcement learning) you have a problem where the "agent" is trained by doing the game over and over. Basically you ascribe a value to every state. Then you assign an expected value of each possible action at a state.
So say you have S states and a actions per state (although you might have more possible moves in one state, and not as many in another), then you want to figure out the most-valuable states from s to be in, and the most valuable actions to take.
In order to figure out the value of states and their corresponding actions, you have to iterate the game through. Probabilistically, a certain sequence of states will lead to victory or defeat, and basically you learn which states lead to failure and are "bad states". You also learn which ones are more likely to lead to victory, and these are subsequently "good" states. They each get a mathematical value associated, usually as an expected reward.
Reward from second-last state to a winning state: +10
Reward if entering a losing state: -10
So the states that give negative rewards then give negative rewards backwards, to the state that called the second-last state, and then the state that called the third-last state and so-on.
Eventually, you have a mapping of expected reward based on which state you're in, and based on which action you take. You eventually find the "optimal" sequence of steps to take. This is often referred to as an optimal policy.
It is true of the converse that normal courses of actions that you are stepping-through while deriving the optimal policy are called simply policies and you are always implementing a certain "policy" with respect to Q-Learning.
Usually the way of determining the reward is the interesting part. Suppose I reward you for each state-transition that does not lead to failure. Then the value of walking all the states until I terminated is however many increments I made, however many state transitions I had.
If certain states are extremely unvaluable, then loss is easy to avoid because almost all bad states are avoided.
However, you don't want to discourage discovery of new, potentially more-efficient paths that don't follow just this-one-works, so you want to reward and punish the agent in such a way as to ensure "victory" or "keeping the pole balanced" or whatever as long as possible, but you don't want to be stuck at local maxima and minima for efficiency if failure is too painful, so no new, unexplored routes will be tried. (Although there are many approaches in addition to this one).
So when you ask "how do you test AI algorithms" the best part is is that the testing itself is how many "algorithms" are constructed. The algorithm is designed to test a certain course-of-action (policy). It's much more complicated than
"turn left every half mile"
it's more like
"turn left every half mile if I have turned right 3 times and then turned left 2 times and had a quarter in my left pocket to pay fare... etc etc"
It's very precise.
So the testing is usually actually how the A.I. is being programmed. Most models are just probabilistic representations of what is probably good and probably bad. Calculating every possible state is easier for computers (we thought!) because they can focus on one task for very long periods of time and how much they remember is exactly how much RAM you have. However, we learn by affecting neurons in a probabilistic manner, which is why the memristor is such a great discovery -- it's just like a neuron!
You should look at Neural Networks, it's mindblowing. The first time I read about making a "brain" out of a matrix of fake-neuron synaptic connections... A brain that can "remember" basically rocked my universe.
A.I. research is mostly probabilistic because we don't know how to make "thinking" we just know how to imitate our own inner learning process of try, try again.
I am running a physics simulation and applying a set of movement instructions to a simulated skeleton. I have a multiple sets of instructions for the skeleton consisting of force application to legs, arms, torso etc. and duration of force applied to their respective bone. Each set of instructions (behavior) is developed by testing its effectiveness performing the desired behavior, and then modifying the behavior with a genetic algorithm with other similar behaviors, and testing it again. The skeleton will have an array behaviors in its set list.
I have fitness functions which test for stability, speed, minimization of entropy and force on joints. The problem is that any given behavior will work for a specific context. One behavior works on flat ground, another works if there is a bump in front of the right foot, another if it's in front of the left, and so on. So the fitness of each behavior varies based on the context. Picking a behavior simply on its previous fitness level won't work because that fitness score doesn't apply to this context.
My question is, how do I program to have the skeleton pick the best behavior for the context? Such as picking the best walking behavior for a randomized bumpy terrain.
In a different answer I've given to this question, I assumed that the "terrain" information you have for your model was very approximate and large-grained, e.g., "smooth and flat", "rough", "rocky", etc. and perhaps only at a grid level. However, if the world model is in fact very detailed, such as from a simulated version of a 3-D laser range scanner, then algorithmic and computational path/motion planning approaches from robotics are likely to be more useful than a machine-learning classifier system.
PATH/MOTION PLANNING METHODS
There are a fairly large number of path and motion planning methods, including some perhaps more suited to walking/locomotion, but a few of the more general ones worth mentioning are:
Visibility graphs
Potential Fields
Sampling-based methods
The general solution approach would be use a path planning method to determine the walking trajectory that your skeleton should follow to avoid obstacles, and then use your GA-based controller to achieve the appropriate motion. This is very much at the core of robotics: sense the world and determine actions and motor control required to achieve some goal(s).
Also, a quick literature search turned up the following papers and a book as a source of ideas and starting points for further investigation. The paper on legged robot motion planning may be especially useful as it discusses several motion planning strategies.
Reading Suggestions
Steven Michael LaValle (2006). Planning Algorithms, Cambridge University Press.
Kris Hauser, Timothy Bretl, Jean-Claude Latombe, Kensuke Harada, Brian Wilcox (2008). "Motion Planning for Legged Robots on Varied Terrain", The International Journal of Robotics Research, Vol. 27, No. 11-12, 1325-1349,
DOI: 10.1177/0278364908098447
Guilherme N. DeSouza and Avinash C. Kak (2002). "Vision for Mobile Robot Navigation: A Survey", IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 24, No. 2, February, pp 237-267.
Why not test the behaviors against a randomized bumpy terrain? Just set the parameters of the GA so that it's a little forgiving, and won't condemn a behavior for one or two failures.
You have two problems:
Bipedal locomotion without senses is very difficult. I've seen good robotic locomotion over rough terrain without senses, but never with only two legs. So the best solution you can possibly find this way might not be very good.
Running a GA is as much art as science. There are a lot of knobs you can turn, and it's hard to find parameters that will allow novelty to grow without drowning it in noise.
Starting simple (e.g. crawling) will help with both of these.
EDIT:
Wait... you're training it over and over on the same randomized terrain? Well no wonder you're having trouble! It's optimizing for that particular layout of rocks and bumps, which is much easier than generalizing. Depending on how your GA works, you might get some benefit from making the course really long, but a better solution is to randomize the terrain for every pass. When it can no longer exploit specific features of the terrain, it will have an evolutionary incentive to generalize. Since this is a more difficult problem it will not learn as quickly as it did before, and it might not be able to get very good at all with its current parameters; be prepared to tinker.
There are three aspects to my answer: (1) control theory, (2) sensing, and (3) merging sensing and action.
CONTROL THEORY
The answer to your problem depends partially on what kind of control scheme you are using: is it feed-forward or feedback control? If the latter, what simulated real-time sensors do you have other than terrain information?
Simply having terrain information and incorporating it into your control strategy would not mean you are using feedback control. It is possible to use such information to select a feed-forward strategy, which seems closest to the problem that you have described.
SENSING
Whether you are using feed-forward or feedback control, you need to represent the terrain information and any other sensory data as an input space for your control system. Part of training your GA-based motion controller should be moving your skeleton through a broad range of random terrain in order to learn feature detectors. The feature detectors classify the terrain scenarios by segmenting the input space into regions critical to deciding what is the best action policy, i.e., what control behavior to employ.
How to best represent the input space depends on the level of granularity of the terrain information you have for your simulation. If it's just a discrete space of terrain type and/or obstacles in some grid space, you may be able to present it directly to your GA without transformation. If, however, the data is in a continuous space such as terrain type and obstacles at arbitrary range/direction, you may need to transform it into a space from which it may be easier to infer spatial relationships, such as coarse-coded range and direction, e.g., near, mid, far and forward, left-forward, left, etc. Gaussian and fuzzy classifiers can be useful for the latter approach, but discrete-valued coding can also work.
MERGING SENSING AND ACTION
Using one of the input-space-encoding approaches above, you have a few options for how to connect behavior selection search space and motion control search space:
Separate the two spaces into two learning problems and use a separate GA to evolve the parameters of a standard multi-layer perceptron neural network. The latter would have your sensor data (perhaps transformed) as inputs and your set of skeleton behaviors as outputs. Instead of using back-propagation or some other ANN-learning method to learn the network weights, your GA could use some fitness function to evolve the parameters over a series of simulated trials, e.g., fitness = distance traveled in a fixed time period toward point B starting from point A. This should evolve over successive generations from completely random selection of behaviors to something more coordinated and useful.
Merge the two search spaces (behavior selection and skeleton motor control) by linking a multi-layer perceptron network as described in (1) above into the existing GA-based controller framework that you have, using the skeleton behavior set as the linkage. The parameter space that will be evolved will be both the neural network weights and whatever your existing controller parameter space is. Assuming that you are using a multi-objective genetic algorithm, such as the NSGA-II algorithm, (since you have multiple fitness functions), the fitness functions would be stability, speed, minimization of entropy, force on joints, etc, plus some fitness function(s) targeted at learning the behavior-selection policy, e.g., distance moved toward point B starting from point A in a fixed time period.
The difference between this approach and (1) above is that you may be able to learn both better coordination of behaviors and finer-grain motor control since the parameter space is likely to be better explored when the two problems are merged as opposed to being separate. The downside is that it may take much longer to converge on reasonable parameter solutions(s), and not all aspects of motor control may be learned as well as they would if the two learning problems were kept separate.
Given that you already have working evolved solutions for the motor control problem, you are probably better off using approach (1) to learn the behavior-selection model with a separate GA. Also, there are many alternatives to the hybrid GA-ANN scheme I described above for learning the latter model, including not learning a model at all and instead using a path planning algorithm as described in a separate answer from me. I simply offered this approach since you are already familiar with GA-based machine learning.
The action selection problem is a robust area of research in both machine learning and autonomous robotics. It's probably well-worth reading up on this topic in itself to gain better perspective and insight into your current problem, and you may be able to devise a simpler strategy than anything I've suggested so far by viewing your problem through the lens of this paradigm.
You're using a genetic algorithm to modify the behavior, so that must mean you have devised a fitness function for each combination of factors. Is that your question?
If yes, the answer depends on what metrics you use to define best walking behavior:
Maximize stability
Maximize speed
Minimize forces on joints
Minimize energy or entropy production
Or do you just try a bunch of parameters, record the values, and then let the genetic algorithm drive you to the best solution?
If each behavior works well in one context and not another, I'd try quantifying how to sense and interpolate between contexts and blend the strategies to see if that would help.
It sounds like at this point you have just a classification problem. You want to map some knowledge about what you are currently walking on to one of a set of classes. Knowing the class of the terrain allows you to then invoke the proper subroutine. Is this correct?
If so, then there are a wide array of classification engines that you can use including neural networks, Bayesian networks, decision trees, nearest neighbor, etc. In order to pick the best fit, we will need more information about your problem.
First, what kind of input or sensory data do you have available to help you identify the behavior class you should invoke? Second, can you describe the circumstances in which you will be training this classifier and what the circumstances are during runtime when you deploy it, such as any limits on computational resources or requirements of robustness to noise?
EDIT: Since you have a fixed number of classes, and you have some parameterized model for generating all possible terrains, I would consider using k-means clustering. The principle is as follows. You cluster a whole bunch of terrains into k different classes, where each cluster is associated with one of your specialized subroutines that performs best for that cluster of terrains. Then when a new terrain comes in, it will probably fall near one of these clusters. You then invoke the corresponding specialized subroutine to navigate that terrain.
Do this offline: Generate enough random terrains to sufficiently sample the parameter space, map these terrains to your sensory space (but remember which points in sensory space correspond to which terrains), and then run k-means clustering on this sensory space corpus where k is the number of classes you want to learn. Your distance function between a class representative C and a point P in sensory space would be simply the fitness function of letting algorithm C navigate the terrain that generated P. You would then get a partitioning of your sensory space into k clusters, each cluster mapping to the best subroutine that you've got. Each cluster will have a representative point in sensory space.
Now during runtime: You will get some unlabeled point in sensory space. Use a different distance function to find the closest representative point to this new incoming point. That tells you what class the terrain is.
Note that the success of this method depends on the quality of the mapping from the parameter space of terrain generation to sensory space, from sensory space to your fitness functions, and the eventual distance function you use to compare points in sensory space.
Note also that if you had enough memory, instead of only using the k representative sensory points to tell you which class an unlabeled sensory point belongs to, you might go through your training set and label all points with the learned class. Then during runtime you pick the nearest neighbor, and conclude that your unlabeled point in sensory space is in the same class as that neighbor.