Path Planning Of Agents Research Articles

Agent path planning based on adaptive polymorphic ant colony optimization

In the path planning of intelligent agents, ant colony algorithm is a popular path solving strategy and has been widely used. However, the traditional ant colony algorithm has problems of local optimum and redundant turning points. The adaptive polymorphic ant colony algorithm greatly improves the search and convergence speed through multi-group division and cooperation mechanism, which helps to enhance the global search ability and avoid falling into the local optimal solution. This paper proposes an improved pheromone update strategy and path selection record table construction to further improve the accuracy of path planning. Finally, the path is processed by cubic B-Spline smoothing curve to effectively reduce the turning points and realize the smoothing of the path. After MATLAB and ROS (Robot Operating System)-Gazebo simulation verification, the results show that the algorithm has good feasibility in complex environments. In summary, the improved adaptive polymorphic ant colony algorithm in this paper brings significant optimization and improvement to the global search of intelligent agents.

Application of BIM Technology in Structural Design of Prefabricated Building Based on Big Data Simulation Modeling Analysis

Aiming at the complex steel bar layout problem in prefabricated building, this research proposes a structural design method of prefabricated building based on big data simulation modeling building information modeling technology. It includes the reinforcement arrangement model based on agent path planning, the intelligent reinforcement arrangement of frame based on artificial potential field method and path optimization, and the modeling of Building information modeling. When analyzing the reinforcement arrangement effect of beam column joints in Prefabricated building, the calculation time of top corner joints is the shortest, 76.8 seconds, while that of middle layer joints is the longest, 141.7 seconds. In direction Y and direction X, differential evolution has the longest calculation time, 27s and 26.57s respectively, while particle swarm optimization algorithm has the shortest calculation time, 2.84s and 3.02s respectively. The algorithm designed through research is significantly superior to other algorithms in terms of computational time. In general, through building information modeling technology and big data simulation modeling, this study realized the collision free layout of rebar in beam column joints of prefabricated building. This improves the speed and efficiency of deepening design, and provides a new solution for structural design of prefabricated building.

Mobile agent path planning under uncertain environment using reinforcement learning and probabilistic model checking

The major challenge in mobile agent path planning, within an uncertain environment, is effectively determining an optimal control model to discover the target location as quickly as possible and evaluating the control system’s reliability. To address this challenge, we introduce a learning-verification integrated mobile agent path planning method to achieve both the effectiveness and the reliability. More specifically, we first propose a modified Q-learning algorithm (a popular reinforcement learning algorithm), called QEA−learning algorithm, to find the best Q-table in the environment. We then determine the location transition probability matrix, and establish a probability model using the assumption that the agent selects a location with a higher Q-value. Secondly, the learnt behaviour of the mobile agent based on QEA−learning algorithm, is formalized as a Discrete-time Markov Chain (DTMC) model. Thirdly, the required reliability requirements of the mobile agent control system are specified using Probabilistic Computation Tree Logic (PCTL). In addition, the DTMC model and the specified properties are taken as the input of the Probabilistic Model Checker PRISM for automatic verification. This is preformed to evaluate and verify the control system’s reliability. Finally, a case study of a mobile agent walking in a grids map is used to illustrate the proposed learning algorithm. Here we have a special focus on the modelling approach demonstrating how PRISM can be used to analyse and evaluate the reliability of the mobile agent control system learnt via the proposed algorithm. The results show that the path identified using the proposed integrated method yields the largest expected reward.

Reinforcement learning based path planning of multiple agents of SwarmItFIX robot for fixturing operation in sheetmetal milling process

SwarmItFIX (self-reconfigurable intelligent swarm fixtures) is a multi-agent setup mainly used as a robotic fixture for large Sheet metal machining operations. A Constraint Satisfaction Problem (CSP) based planning model is utilized currently for computing the locomotion sequence of multiple agents of the SwarmItFIX. But the SwarmItFIX faces several challenges with the current planner as it fails on several occasions. Moreover, the current planner computes only the goal positions of the base agent, not the path. To overcome these issues, a novel hierarchical planner is proposed, which employs Monte Carlo and SARSA TD based model-free Reinforcement Learning (RL) algorithms for the computation of locomotion sequences of head and base agents, respectively. These methods hold two distinct features when compared with the existing methods (i) the transition model is not required for obtaining the locomotion sequence of the computational agent, and (ii) the state-space of the computational agent become scalable. The obtained results show that the proposed planner is capable of delivering optimal makespan for effective fixturing during the sheet metal milling process.

Analytic Velocity Obstacle for Efficient Collision Avoidance Computation and a Comparison Study With Sampling and Optimization-Based Approaches

Abstract Velocity obstacle (VO) is one of the popular reactive navigation algorithms for the path planning of autonomous agents. The collision-free property can be guaranteed if the agent is able to choose a velocity outside the VO region under the assumption that obstacles maintain a constant velocity within the control cycle time of the agent. To date, the selection of the optimal velocity relies on either sampling or optimization approaches. The sampling approach can maintain the same amount of computation cost but may miss feasible solutions under collision risks with an insufficient number of samples. The optimization approach such as the linear programming demands convexity of the constraints in the velocity space which may not be satisfied considering non-holonomic agents. In addition, the algorithm has varying computation demands depending on the navigation situation. This paper proposes an analytic approach for choosing a candidate velocity rather than relying on sampling or optimization approaches. The analytic approach can significantly reduce computation costs without sacrificing performance. Agents with both holonomic and non-holonomic constraints are considered to demonstrate the performance and efficiency of the proposed approach. Extensive comparison studies with static, non-reactive, and reactive moving obstacles demonstrate that the analytical VO is computationally much more efficient than the optimization-based approach and performs better than the sampling-based approach. Major video results of this paper can be accessed online.

The Shortest Path Planning Based on Reinforcement Learning

This paper proposes a shortest path planning of agent in an environment based on reinforcement learning. This method adopts the Q-learning algorithm, which has gained increasingly using in agent path planning recently. This algorithm can be fully designed to a reasonable environment model and applied to other professional fields. Yet, reinforcement learning in path planning still needs to improve the rate of convergence. The learning process takes several iterations and spends a long time to find the final path. A local optimal problem in search process deals with other explorations that provided shortest path for agent to take optimal actions. For the agent, the reward value of inclined movement is introduced to find the shortest path. The autonomous obstacle avoidance is utilized to obtain the optimal path in the process. Finally, the model of intelligent agent movement environment has established in matlab. The effectiveness of the algorithm has proved by simulation results.

Hexagon-Based Generalized Voronoi Diagrams Generation for Path Planning of Intelligent Agents

Grid-based Generalized Voronoi Diagrams (GVDs) are widely used to represent the surrounding environment of intelligent agents in the fields of robotics, computer games, and military simulations, which improve the efficiency of path planning of intelligent agents. Current studies mainly focus on square-grid-based GVD construction approaches, and little attention has been paid to constructing GVDs from hexagonal grids. In this paper, an algorithm named hexagon-based crystal growth (HCG) is presented to extract GVDs from hexagonal grids. In addition, two thinning patterns for obtaining one-cell-wide GVDs from rough hexagon-based GVDs are proposed. On the basis of the principles of a leading square-grid-based algorithm named Brushfire, a hexagon-based Brushfire algorithm is realized. A comparison of the HCG and the hexagon-based Brushfire algorithm shows that HCG is much more efficient. Further, the usefulness of hexagon-based GVDs for the path planning tasks of intelligent agents is demonstrated using several representative simulation experiments.

Artificial intelligence planners for multi-head path planning of SwarmItFIX agents

Sheet metal manufacturing is finding wide applications in automotive and aerospace industries. Handling of giant sheet materials in manufacturing industries is one of the key problems. Utilization of robots, viz SwarmItFIX, will address this problem and automate the fixturing process, which greatly reduces lead time and thus the production cost. Implementation of intelligence into the robots will further improve efficiency in handling and reduce manufacturing inaccuracies. In this work, two different novel planners are proposed which do path planning for the heads of the SwarmItFIX agents. The environment of the problem is modeled as a Markov Decision Problem. The first planner uses the Value Iteration and Policy Iteration (PI) algorithms individually and the second planner performs the Monte Carlo control reinforcement learning. Finally, when the simulation is done and parameters of the proposed three algorithms along with existing Constraint Satisfaction Problem algorithm are compared with each other. It is observed that the proposed PI algorithm returns the plan much faster than the other algorithms. In the near future, the efficient planning model will be tested and implemented into the SwarmItFIX setup at the PMAR laboratory, University of Genoa, Italy.

Consensus-Based Odor Source Localization by Multiagent Systems.

This paper presents an investigation of the task of localizing an unknown source of an odor by heterogeneous multiagent systems. A hierarchical cooperative control strategy has been proposed as a potential candidate to solve the problem. The agents are driven into consensus as soon as the information about the location of source is acquired. The controller has been designed in a hierarchical manner of group decision making, agent path planning, and robust control. In group decision making, the particle swarm optimization algorithm has been used along with the information of the movement of odor molecules to predict the odor source location. Next, a trajectory has been mapped using this predicted location of source, and the information is passed to the control layer. A variable structure control has been used in the control layer due to its inherent robustness and disturbance rejection capabilities. Cases of movement of agents toward the source under consensus and parallel formation have been discussed. The efficacy of the proposed scheme has been confirmed by simulations.

Coordinated Selection and Timing of Multiple Trajectories of Discretely Mobile Robots

The paper addresses the multi-agent path planning (MPP) of mobile agents with multiple goals taking into consideration the kinematic constraints of each agent. The “Swing and Dock” (SaD) robotic system being discussed uses discrete locomotion, where agents swing around fixed pins and dock with their mounting legs to realize displacement from one point to another. The system was developed as a subsystem for mobile robotic fixture (SwarmItFix). Previous work dealt with MPP for SaD agents using the concept of extended temporal graph with Integer Linear Programming (ILP) based formulations. The approach discretized time into unit steps, whereas in reality, the agents are constrained by velocity limits. Hence, a real-time schedule is required to accurately plan the agent movement in a working scenario. We utilize the concept of simple temporal network and extend our ILP formulations to model the velocity kinematic constraints. The mathematical formulations are implemented and tested using a GUROBI solver. Computational results display the effectiveness of the approach.

Plan Restructuring in Multi Agent Planning

In this paper we consider a multi agent planning problem where given a set of plans of the individual agents that may lead to conflict during execution, the task is to find a conflict-free plan for the agents. For this we give a formal model of the multi agent planning problem (MAPP) and give a description of the multi agent path planning. We demonstrate this technique in a grid world domain. The experimental results show promise for our method.

Path Planning Problem

Path planning is the way of determination of a collision free path between start and goal position through obstacles cluttered in a workspace. Though it is a complex problem, but it is an essential task for the navigation and controlling the motion of autonomous robot manipulators. This NP-complete problem (those problems is difficult to solve specially in a dynamic environment where the optimal path needs to be rerouted in real time when a new obstacle appears. This paper provides two categories of path planning approaches:Deterministic and Probabilistic approaches. Deterministic methods allow achieving the same result in each execution with the same initial conditions. They are perfectly predictable, hence suitable for static environment, but not effective when they are used in a real time environment as there could be sudden changes in environment. The most used solution to overcome the problem of real time environment are the probabilistic methods such as Particle swarm optimization[pso], Ant colony optimization[aco], genetic algorithm[ga], multi agent path planning,etc. General Terms Deterministic algorithm, probabilistic algorithm, aco, pso, ga, A*algorithm, dijkstra, multi agent path planning.

Steering Behaviour for Computer based Intelligent Agents

paper presents steering behaviours through intelligent agents populating virtual or simulated environments. We present a reliable effort towards automatic path planning of intelligent computer agents with the integration of trajectory optimisation and obstacle avoidance techniques. Our approach will increase the level of accuracy in movement and would be reliable in the sense that the best computable path will always be available. Behaviour programming is facilitated by a set of presentations of the environment that uses convenient frames of reference in two dimensional coordinate systems. Four different steering ways are examined for movement of intelligent agents. Combinations of steering behaviours can be used to achieve high level goals of movement or locomotion. Our approach can be tailored to suit any particular need which requires automation in locomotion. Keywordsual environments, intelligent agent, steering behaviour

Multiagent Information Fusion and Cooperative Control in Target Search

This paper addresses cooperative search for multiple stationary ground targets by a group of unmanned aerial vehicles with limited sensing and communication capabilities. The whole surveillance region is partitioned into cells where each cell is associated with a probability of target existence within the cell, which constitutes a probability map for the whole region. Each agent keeps an individual probability map and updates the map individually with measurements according to Bayesian rule. A nonlinear transformation of the probability map is introduced to simplify the computation by linearizing the Bayesian update. A consensus-like distributed fusion scheme is proposed for multiagent map fusion. We prove that all the individual probability maps converge to the same one that reflects the true existence or nonexistence of targets within each cell. Coverage and topology control algorithms are designed for the path planning of mobile agents. Moreover, the performance of the fusion scheme for asynchronous implementations of sampling and communication is analyzed. Finally, the effectiveness of the proposed algorithms is illustrated via simulations.

Artificial Intelligence in the Path Planning Optimization of Mobile Agent Navigation

Many difficult problem solving require computational intelligence. One of the major directions in artificial intelligence consists in the development of efficient computational intelligence algorithms, like: evolutionary algorithms, and neural networks. Systems, that operate in isolation or cooperate with each other, like mobile robots could use computational intelligence algorithms for different problems/tasks solving, however in their behavior could emerge an intelligence called system's intelligence, intelligence of a system. The traveling salesman problem TSP has a large application area. It is a well-known business problem. Maximum benefits TSP, price collecting TSP have a large number of economic applications. TSP is also used in the transport logic Raja, 2012. It also has a wide range of applicability in the mobile robotic agent path planning optimization. In this paper a mobile robotic agent's path planning will be discussed, using unsupervised neural networks for the TSP solving, and from the TSP results the finding of a closely optimal path between two points in the agent's working area. In the paper a modification of the criteria function of the winner neuron selection will also be presented. At the end of the paper measurement results will be presented.

Parallel Evaluation of a Spatial Traversability Cost Function on GPU for Efficient Path Planning

A parallel version of the traditional grid based cost-to-go function generation algorithm used in robot path planning is introduced. The process takes advantage of the spatial layout of an occupancy grid by concurrently calculating the next wave front of grid cells usually evaluated sequentially in traditional dynamic programming algorithms. The algorithm offers an order of magnitude increase in run time for highly obstacle dense worst-case environments. Efficient path planning of real world agents can greatly increase their accuracy and responsiveness. The process and theoretical analysis are covered before the results of practical testing are discussed.

Real-Time Path Planning in Dynamic Virtual Environments Using Multiagent Navigation Graphs

We present a novel approach for efficient path planning and navigation of multiple virtual agents in complex dynamic scenes. We introduce a new data structure, Multi-agent Navigation Graph (MaNG), which is constructed using first- and second-order Voronoi diagrams. The MaNG is used to perform route planning and proximity computations for each agent in real time. Moreover, we use the path information and proximity relationships for local dynamics computation of each agent by extending a social force model [Helbing05]. We compute the MaNG using graphics hardware and present culling techniques to accelerate the computation. We also address undersampling issues and present techniques to improve the accuracy of our algorithm. Our algorithm is used for real-time multi-agent planning in pursuit-evasion, terrain exploration and crowd simulation scenarios consisting of hundreds of moving agents, each with a distinct goal.

A low cost vision based localization system using fiducial markers

This paper investigates a mobile robot self-localization system based on fiducial markers which are placed on the ceiling. Recently there have been many articles related to path planning and coordination of mobile agents within an unknown environment. These algorithms demand dedicated (and often expensive) hardware that is capable of handling complex tasks in real time. In this paper, we have explored the possibility to realize an inexpensive and simple navigation system, based on passive fiducial markers, which are able to guide an autonomous mobile robot along a predefined path. Fiducial markers provide not only improved performance in runtime, but also much better identification and localization. Marker design based on circle shape is presented. Using a low-cost webcam and appropriate marker detection algorithm three features are determined: robot position, new movement direction and marker ID. Experimental results are presented at the end of the paper.