A solution to the path-planning problem for mobile robots is proposed. The solution is based on a geometrical model, a logical model, and an associative memory. The geometrical modeling is done by a graph produced by ...
详细信息
A solution to the path-planning problem for mobile robots is proposed. The solution is based on a geometrical model, a logical model, and an associative memory. The geometrical modeling is done by a graph produced by analog-mapping preference gangways in the environment. The path search is made by a best-first algorithm in the logical model. The costs of the individual graph arcs are dynamically updated. During the search, the associative memory is repeatedly consulted for an available path. When a valid path is found it is learned by the memory for later use. For the memory, the path is tristate-encoded and preprocessed by the graph connection matrix. An example is given to illustrate the operation of the associative memory.< >
Algorithms are presented for continuous-time quadratic optimization of motion control. Explicit solutions to the Hamilton-Jacobi equation for optimal control of rigid-body motion are found by solving an algebraic matr...
详细信息
Algorithms are presented for continuous-time quadratic optimization of motion control. Explicit solutions to the Hamilton-Jacobi equation for optimal control of rigid-body motion are found by solving an algebraic matrix equation. The system stability is investigated according to Lyapunov function theory, and it is shown that global asymptotic stability holds. It is also shown how optimal control and adaptive control may act in concert in the case of unknown or uncertain system parameters. The solution results in natural design parameters in the form of square weighting matrices as known from linear quadratic optimal control. The proposed optimal control is useful for motion control, trajectory planning, and motion analysis.< >
We describe how to determine the time delays caused by synchronous and asynchronous control loops. These time delays can vary in a quite irregular way and can cause problems for the control loops closed over a communi...
详细信息
We describe how to determine the time delays caused by synchronous and asynchronous control loops. These time delays can vary in a quite irregular way and can cause problems for the control loops closed over a communication network of, for instance, field-buses. Different setups allowing an arbitrary number of communication layers are analyzed. The theoretical derivations are illustrated with a simple but realistic example.
In this paper, the heuristic prediction method (HPM) is extended to the long time horizon constrained optimal control problems. The key features of this method are that a heuristic term is added to each of the subprob...
详细信息
ISBN:
(纸本)0780300424
In this paper, the heuristic prediction method (HPM) is extended to the long time horizon constrained optimal control problems. The key features of this method are that a heuristic term is added to each of the subproblems, and that a prediction method is adopted in the high level problem. Numerical results show its feasibility and its potential applications to real-time control problems.< >
Reinforcement learning (RL) control provides a means to deal with uncertainty and nonlinearity associated with control tasks in an optimal way. The class of actorcritic RL algorithms proved useful for controlsystems ...
详细信息
ISBN:
(纸本)9781467320658
Reinforcement learning (RL) control provides a means to deal with uncertainty and nonlinearity associated with control tasks in an optimal way. The class of actorcritic RL algorithms proved useful for controlsystems with continuous state and input variables. In the literature, model-based actor-critic algorithms have recently been introduced to considerably speed up the learning by constructing online a model through local linear regression (LLR). It has not been analyzed yet whether the speed-up is due to the model learning structure or the LLR approximator. Therefore, in this paper we generalize the model learning actor-critic algorithms to make them suitable for use with an arbitrary function approximator. Furthermore, we present the results of an extensive analysis through numerical simulations of a typical nonlinear motion control problem. The LLR approximator is compared with radial basis functions (RBFs) in terms of the initial convergence rate and in terms of the final performance obtained. The results show that LLR-based actor-critic RL outperforms the RBF counterpart: it gives quick initial learning and comparable or even superior final control performance.
We consider controlsystems with switchings in dynamics and control objective in the form of fixed terminal state. Two types of parameter deviations are discussed. In the first case, infinitesimal changes of system pa...
详细信息
We consider controlsystems with switchings in dynamics and control objective in the form of fixed terminal state. Two types of parameter deviations are discussed. In the first case, infinitesimal changes of system parameters are assumed, and in the second one, small changes in switching time are considered. Conditions for reachability of the control objective are formulated.
A new method for solving hierarchical optimal control problems is proposed in this paper which combines the accelerated simulated annealing algorithm and the constrained optimal control neural network in an appropriat...
详细信息
A new method for solving hierarchical optimal control problems is proposed in this paper which combines the accelerated simulated annealing algorithm and the constrained optimal control neural network in an appropriate manner. Two different forms of explicitly defined high-level constraints of the hierarchical problems are treated by the different methods. To cope with the high-level constraints depending only on the high-level control variables, an auxiliary problem is introduced at the upper level, avoiding using the penalty function method to deal with such constraints in an inefficient way. The computational results for numerical examples demonstrate the validity and better performance of the proposed method in terms of solution quality and computational efficiency.
It was demonstrated that the standard selection of input weights and biases for Extreme Learning Machine (ELM) may lead to ill-conditioning of the output weights calculation and result in great values of the output we...
详细信息
ISBN:
(纸本)9781509039104
It was demonstrated that the standard selection of input weights and biases for Extreme Learning Machine (ELM) may lead to ill-conditioning of the output weights calculation and result in great values of the output weights. Two slight modifications of the standard approach were proposed: i) addition of some distance-based neurons (like Radial Basis Functions - RBF) with centers situated at numerically identified local extrema of the training data, ii) modification of random generation of input weights and biases for sigmoid neurons to enable enhanced variation of activation functions. The first modification improves the modeling accuracy for a moderate number of hidden neurons and the second allows to improve the conditioning of output weights calculation and reduce the optimal output weights significantly.
Heterogeneity in cell populations is a major factor in the dynamics of cellular systems in living tissue or microbial colonies. This heterogeneity needs to be taken into account for the interpretation of experimental ...
详细信息
This paper mainly studied the algorithm for compressive sensing signal reconstruction, aiming to reduce the time of reconstruction algorithm. By analyzing the characteristics of the commonly used of orthogonal matchin...
详细信息
This paper mainly studied the algorithm for compressive sensing signal reconstruction, aiming to reduce the time of reconstruction algorithm. By analyzing the characteristics of the commonly used of orthogonal matching pursuit, we choosed two key steps of it to improve. Using stagewise algorithm in the process of atomic selection, while using gradient pursuit algorithm in the process of projecting the allowance to the subspace, to achieve a reconstruction algorithm, which based on stagewise gradient pursuit. The simulation results show that, compared with basic algorithm based on the orthogonal matching pursuit, the proposed algorithm has obvious reduction in time of reconstruction.
暂无评论