Guan-Chun Luh

MOIA: Multi-objective immune algorithm

The paper describes a novel algorithm for finding Pareto optimal solutions to multi-objective optimization problems based on the features of a biological immune system. Inter-relationships within the proposed multi-objective immune algorithm (MOIA) resemble antibody-antigen relationships in terms of specificity, germinal center, and the memory characteristics of adaptive immune responses. Gene fragment recombination and several antibody diversification schemes (including somatic recombination, somatic mutation, gene conversion, gene reversion, gene drift, and nucleotide addition) were incorporated into the MOIA in order to improve the balance between exploitation and exploration. Using five performance metrics, MOIA simulation figures were compared with data derived from a strength Pareto evolutionary algorithm (SPEA). The results indicate that the MOIA outperformed the SPEA in several areas.

A multi-modal immune algorithm for the job-shop scheduling problem

This paper describes the application of an artificial immune system to a scheduling application. A novel approach multi-modal immune algorithm is proposed for finding optimal solutions to job-shop scheduling problems emulating the features of a biological immune system. Inter-relationships within the proposed algorithm resemble antibody molecule structure, antibody-antigen relationships in terms of specificity, clonal proliferation, germinal center, and the memory characteristics of adaptive immune responses. Gene fragment recombination and several antibody diversification schemes including somatic recombination, somatic mutation, gene conversion, gene reversion, gene drift, and nucleotide addition were incorporated into the algorithm in order to improve the balance between exploitation and exploration. In addition, niche antibody was employed to discover multi-modal solutions. Numerous well-studied benchmark examples in job-shop scheduling problems were utilized to evaluate the proposed approach. The results indicate the effectiveness and flexibility of the immune algorithm.

Nonlinear parity equation based residual generation for diagnosis of automotive engine faults

Abstract The parity equation residual generation method is a model-based fault detection and isolation scheme that has been applied with some success to the problem of monitoring the health of engineering systems. However, this scheme fails when applied to significantly nonlinear systems. This paper presents the application of a nonlinear parity equation residual generation scheme that uses forward and inverse dynamic models of nonlinear systems, to the problem of diagnosing sensor and actuator faults in an internal combustion engine, during execution of the United States Environmental Protection Agency Inspection and Maintenance 240 driving cycle. The Nonlinear AutoRegressive Moving Average Model with eXogenous inputs technique is used to identify the engine models required for residual generation. The proposed diagnostic scheme is validated experimentally and is shown to be sensitive to a number of input and sensor faults while remaining robust to the unmeasured load torque disturbance.

Structural topology optimization using ant colony optimization algorithm

The ant colony optimization (ACO) algorithm, a relatively recent bio-inspired approach to solve combinatorial optimization problems mimicking the behavior of real ant colonies, is applied to problems of continuum structural topology design. An overview of the ACO algorithm is first described. A discretized topology design representation and the method for mapping ants trail into this representation are then detailed. Subsequently, a modified ACO algorithm with elitist ants, niche strategy and memory of multiple colonies is illustrated. Several well-studied examples from structural topology optimization problems of minimum weight and minimum compliance are used to demonstrate its efficiency and versatility. The results indicate the effectiveness of the proposed algorithm and its ability to find families of multi-modal optimal design.

A binary particle swarm optimization for continuum structural topology optimization

The particle swarm optimization (PSO) algorithm, a relatively recent bio-inspired approach to solve combinatorial optimization problems mimicking the social behavior of birds flocking, is applied to problems of continuum structural topology design for the purpose of investigating optimal topologies and automatically creating innovative solutions. An overview of the PSO and binary PSO algorithms are first described. A discretized topology design representation and the method for mapping binary particle into this representation are then detailed. Subsequently, a modified binary PSO algorithm that adopts the concept of genotype-phenotype representation is illustrated. Several well-studied examples from structural topology optimization problems of minimum weight and minimum compliance are used to demonstrate its efficiency and versatility. The results indicate the effectiveness of the proposed algorithm and its ability to find families of structural topologies.

An immunological approach to mobile robot reactive navigation

In this paper, a reactive immune network (RIN) is proposed and employed for mobile robot navigation within unknown environments. Rather than building a detailed mathematical model of artificial immune systems, this study tries to explore the principle in an immune network focusing on its self-organization, adaptive learning capability, and immune feedback. In addition, an adaptive virtual target method is integrated to solve the local minima problem in navigation. Several trapping situations designed by the early researchers are adopted to evaluate the performance of the proposed architecture. Simulation results show that the mobile robot is capable of avoiding obstacles, escaping traps, and reaching the goal efficiently and effectively.

Reactive Immune Network Based Mobile Robot Navigation

In this paper, a Reactive Immune Network (RIN) is proposed and applied to intelligent mobile robot learning navigation strategies within unknown environments. Rather than building a detailed mathematical model of immune systems, we try to explore the principle in immune network focusing on its self-organization, adaptive learning capability and immune memory. Modified virtual target method is integrated to solve local minima problem. Several trap situations designed by early researchers are employed to evaluate the performance of the proposed immunized architecture. Simulation results show that the robot is capable to avoid obstacles, escape traps, and reach goal effectively.

Immune model-based fault diagnosis

In this paper, a novel approach to immune model-based fault diagnosis methodology for nonlinear systems is presented. The diagnosis scheme consists of forward/inverse immune model identification, filtered residual generation, the fault alarm concentration (FAC), and the artificial immune regulation (AIR). A two-link manipulator simulation was employed to validate the effectiveness and robustness of the diagnosis approach. The simulation results show that it can detect and isolate actuator faults, sensor faults, and system component faults efficiently.

PCA based immune networks for human face recognition

This paper proposes a face recognition method using artificial immune networks based on principal component analysis (PCA). The PCA abstracts principal eigenvectors of the image in order to get best feature description, hence to reduce the number of inputs of immune networks. Henceforth these image data of reduced dimensions are input into immune network classifiers to be trained. Subsequently the antibodies of the immune networks are optimized using genetic algorithms. The performance of the present method was evaluated employing the AT&T Laboratories Cambridge database. The results show that this method gains higher recognition rate in contrast with most of the developed methods.

Behavior-based intelligent mobile robot using an immunized reinforcement adaptive learning mechanism

Abstract In this paper, a novel immunized reinforcement adaptive learning mechanism employing a behavior-based knowledge and the on-line adapting capabilities of the immune system is proposed and applied to an intelligent mobile robot. Rather than building a detailed mathematical model of immune systems, we try to explore principles in the immune system focusing on its self-organization, adaptive capability and immune memory. Two levels of the immune system, underlying the ‘micro’ level of cell interactions, and emergent ‘macro’ level of the behavior of the system are investigated. To evaluate the proposed immunized architecture, a ‘food foraging work’ simulation environment containing a mobile robot, foods, with/without obstacles is created to simulate the real world. The simulation results validate several significant characteristics of the immunized architecture: adaptability, learning, self-organizing, and stable ecological niche approaching.