Dunwei Gong

Robot path planning in uncertain environment using multi-objective particle swarm optimization

In many real-world applications, workspace of robots often involves various danger sources that robots must evade, such as fire in rescue mission, landmines and enemies in war field. Since it is either impossible or too expensive to get their precise positions, decision-makers know only their action ranges in most cases. This paper proposes a multi-objective path planning algorithm based on particle swarm optimization for robot navigation in such an environment. First, a membership function is defined to evaluate the risk degree of path. Considering two performance merits: the risk degree and the distance of path, the path planning problem with uncertain danger sources is described as a constrained bi-objective optimization problem with uncertain coefficients. Then, a constrained multi-objective particle swarm optimization is developed to tackle this problem. Several new operations/improvements such as the particle update method based on random sampling and uniform mutation, the infeasible archive, the constrained domination relationship based on collision times with obstacles, are incorporated into the proposed algorithm to improve its effectiveness. Finally, simulation results demonstrate the capability of our method to generate high-quality Pareto optimal paths.

Multi-objective Particle Swarm Optimization for Robot Path Planning in Environment with Danger Sources

Aiming at robot path planning in an environment with danger sources, a global path planning approach based on multi-objective particle swarm optimization is presented in this paper. First, based on the environment map of a mobile robot described with a series of horizontal and vertical lines, an optimization model of the above problem including two indices, i.e. the length and the danger degree of a path, is established. Then, an improved multi-objective particle swarm optimization algorithm of solving the above model is developed. In this algorithm, a self-adaptive mutation operation based on the degree of a path blocked by obstacles is designed to improve the feasibility of a new path. To improve the performance of our algorithm in exploration, another archive is adopted to save infeasible solutions besides a feasible solutions archive, and the global leader of particles is selected from either the feasible solutions archive or the infeasible one. Moreover, a constrained Pareto domination based on the degree of a path blocked by obstacles is employed to update local leaders of a particle and the two archives. Finally, simulation results confirm the effectiveness of our algorithm.

Multi-Objective Particle Swarm Optimization Approach for Cost-Based Feature Selection in Classification

Feature selection is an important data-preprocessing technique in classification problems such as bioinformatics and signal processing. Generally, there are some situations where a user is interested in not only maximizing the classification performance but also minimizing the cost that may be associated with features. This kind of problem is called cost-based feature selection. However, most existing feature selection approaches treat this task as a single-objective optimization problem. This paper presents the first study of multi-objective particle swarm optimization (PSO) for cost-based feature selection problems. The task of this paper is to generate a Pareto front of nondominated solutions, that is, feature subsets, to meet different requirements of decision-makers in real-world applications. In order to enhance the search capability of the proposed algorithm, a probability-based encoding technology and an effective hybrid operator, together with the ideas of the crowding distance, the external archive, and the Pareto domination relationship, are applied to PSO. The proposed PSO-based multi-objective feature selection algorithm is compared with several multi-objective feature selection algorithms on five benchmark datasets. Experimental results show that the proposed algorithm can automatically evolve a set of nondominated solutions, and it is a highly competitive feature selection method for solving cost-based feature selection problems.

A niching PSO-based multi-robot cooperation method for localizing odor sources

Aiming at the problem of multiple odor sources localization, a multi-robot cooperation method based on niching particle swarm optimization is presented in this study. In this method, a robot is regarded as a particle, particles located at a neighbor form a niche, and different niches are employed to localize different odor sources synchronously. In order to localize more odor sources, the size of each niche is dynamically adjusted based on the aggregation degree of its elements. A niche merging strategy, based on the similarity of optimal particles found by niches, is proposed to prevent particles repeatedly searching for the same region. In addition, some real conditions such as the sampling/recovery time of a sensor and the velocity limit of a robot are considered when updating the position of a particle. Finally, the proposed method is applied to various scenarios of localizing multiple odor sources, and the experimental results confirm its effectiveness.

Multi-objective particle swarm optimization based on minimal particle angle

Particle swarm optimization is a computational intelligence method of solving the multiobjective optimization problems. But for a given particle, there is no effective way to select its globally optimal particle and locally optimal particle. The particle angle is defined by the particles objective vector. The globally optimal particle is selected according to the minimal particle angle. Updating the locally optimal particle and particle swarm is based on the Pareto dominance relationship between the locally optimal particle and the offspring particles and the particles density. A multiobjective particle swarm optimization based on the minimal particle angle is proposed. The algorithm proposed is compared with sigma method ,NSPSO method and NSGA-II method on four complicated benchmark multiobjective function optimization problems. It is shown from the results that the Pareto front obtained with the algorithm proposed in this paper has good distribution, approach and extension properties.

Modified particle swarm optimization for odor source localization of multi-robot

Odor source localization is very important in real-world applications. We studied the problem of odor source localization and presented a modified particle swarm optimization algorithm for odor source localization of multi-robot. The algorithm dynamically adjusts two learning factors in the velocity update equation based on the effect of wind on self-cognition and social cognition of a particle. In addition, an artificial potential field method is employed to improve the performance of our algorithm. We conducted various experiments in time-varying environments, and the experimental results confirm the superiority of our algorithm.

Interactive genetic algorithms with individual's fuzzy fitness

Interactive genetic algorithms are effective methods to solve an optimization problem with implicit or fuzzy indices, and have been successfully applied to many real-world optimization problems in recent years. In traditional interactive genetic algorithms, many researchers adopt an accurate number to express an individuals fitness assigned by a user. But it is difficult for this expression to reasonably reflect a users fuzzy and gradual cognitive to an individual. We present an interactive genetic algorithm with an individuals fuzzy fitness in this paper. Firstly, we adopt a fuzzy number described with a Gaussian membership function to express an individuals fitness. Then, in order to compare different individuals, we generate a fitness interval based on @a-cut set, and obtain the probability of individual dominance by use of the probability of interval dominance. Finally, we determine the superior individual in tournament selection with size two based on the probability of individual dominance, and perform the subsequent evolutions. We apply the proposed algorithm to a fashion evolutionary design system, a typical optimization problem with an implicit index, and compare it with two interactive genetic algorithms, i.e., an interactive genetic algorithm with an individuals accurate fitness and an interactive genetic algorithm with an individuals interval fitness. The experimental results show that the proposed algorithm is advantageous in alleviating user fatigue and looking for users satisfactory individuals.

Interactive genetic algorithm assisted with collective intelligence from group decision making

Interactive genetic algorithms (IGAs) have been successfully applied to optimize problems with aesthetic criteria by embedding the intelligent evaluations of a user into the evolutionary process. User fatigue caused by frequent interactions, however, often greatly impairs the potentials of IGAs on solving complicated optimization problems. Taking the benefits of collective intelligence into account, we here present an IGA with collective intelligence which is derived from a mechanism of group decision making. An IGA with interval individual fitness is focused here and it can be separately conducted by multiple users at the same time. The collective intelligence of all participated users, represented with social and individual knowledge, is first collected by using a modified group decision making method. Then the strategy of applying the collective intelligence to initialize and guide the single evolution of the IGA is given. With such a multi-user promoted IGA framework, the performance of a single IGA is expected to be evidently improved. In a local network environment, the algorithm is applied to a fashion design system and the results empirically demonstrate that the algorithm can not only alleviate user fatigue but also increase the opportunities of IGAs on finding most satisfactory solutions.

Evolutionary generation of test data for path coverage with faults detection

The aim of software testing is to find faults in the program under test. Previous methods of path-oriented test data generation can generate test data traversing target paths, but they may not guarantee to find faults in the program. We present a method of evolutionary generation of test data for path coverage with faults detection in this paper. First, we establish a mathematical model of the problem considered in this paper, in which the number of faults detected in the path traversed by test data, and the risk level of faults are optimization objectives, and the approach level of the traversed path from the target one is a constraint. Then, we generate test data using a multi-objective evolutionary optimization algorithm with constraints. Finally, we apply the proposed method in a benchmark program bubble sort and an industrial program totinfo, and compare it with the traditional method. The experimental results conform that our method can generate test data that not only traverse the target path but also detect faults in it. Our achievement provides a novel way to generate test data for path coverage with faults detection.

A Novel Multi-agent Based Complex Process Control System and Its Application

Complex process control systems need a hybrid control mode, which combines hierarchical structure with decentralized control units. Autonomy of agents and cooperation capability between agents in multi-agent system provide basis for realization of the hybrid control mode. A novel multi-agent based complex process control system is proposed. Semantic representation of a control-agent is presented utilizing agent-oriented programming. A novel temporal logic analysis of a control-agent is proposed using Petri nets. Collaboration relationships among control-agents are analyzed based on extended contract net protocol aiming at the lack of reference [1]. Taken pressure control of recycled gas with complicated disturbances as an application, five kinds of control-agents are derived from control-agent. Reachable marking tree and different transition of each derived control-agent are analyzed in detail. Actual running effect indicates multi-agent based hybrid control mode is rationality and flexible. Temporal logic analysis based on Petri nets ensures the reachability of the systems. Extended contract net protocol provides a reasonable realization for collaboration relationships.