This paper presents educational software developed for designing FIR and IIR digital filters using two evolutionary algorithms (EAs); namely immune algorithms (IAs) and genetic algorithms (GAs),together with quasi-Newton line search algorithm (QNLS). This software provides the user the ability to design one- and two-dimensional; low-pass, high-pass, band-pass and band-stop digital filters with arbitrary magnitude and group delay specifications. The software is evaluated by making the assessment quizzes for electrical engineering students and instructors. Students’ responses are very positive. A number of recommendations are made in this work based on instructor observation and students’ evaluations.

This paper presents educational software developed for designing FIR and IIR digital filters using two evolutionary algorithms (EAs); namely immune algorithms (IAs) and genetic algorithms (GAs),together with quasi-Newton line search algorithm (QNLS). This software provides the user the ability to design one- and two-dimensional; low-pass, high-pass, band-pass and band-stop digital filters with arbitrary magnitude and group delay specifications. The software is evaluated by making the assessment quizzes for electrical engineering students and instructors. Students’ responses are very positive. A number of recommendations are made in this work based on instructor observation and students’ evaluations.

This paper presents educational software developed for designing FIR and IIR digital filters using two evolutionary algorithms (EAs); namely immune algorithms (IAs) and genetic algorithms (GAs),together with quasi-Newton line search algorithm (QNLS). This software provides the user the ability to design one- and two-dimensional; low-pass, high-pass, band-pass and band-stop digital filters with arbitrary magnitude and group delay specifications. The software is evaluated by making the assessment quizzes for electrical engineering students and instructors. Students’ responses are very positive. A number of recommendations are made in this work based on instructor observation and students’ evaluations.

Despite the considerable amount of research related to immune algorithms and it applications in numerical optimization, digital filters design, and data mining, there is still little work related to issues as important as sensitivity analysis, [1][4]. Other aspects, such as convergence speed and parameters adaptation, have been practically disregarded in the current specialized literature [7][8]. The convergence speed of the immune algorithm heavily depends on its main control parameters: population size, replication rate, mutation rate, clonal rate and hypermutation rate. In this paper we investigate the effect of control parameters variation on the convergence speed for single and multiobjective optimization problems. Three examples are devoted for this purpose; namely the design of 2D recursive digital filter, minimization of simple function, and banana function. The effect of each parameter on the convergence speed of the IA is studied considering the other parameters with fixed values and taking the average of 100 times independent runs. Then, the concluded rules are applied on some examples introduced in [2] and [3]. Computational results show how to select the immune algorithm parameters to speedup the algorithm convergence and to obtain the optimal solution.

Despite the considerable amount of research related to immune algorithms and it applications in numerical optimization, digital filters design, and data mining, there is still little work related to issues as important as sensitivity analysis, [1][4]. Other aspects, such as convergence speed and parameters adaptation, have been practically disregarded in the current specialized literature [7][8]. The convergence speed of the immune algorithm heavily depends on its main control parameters: population size, replication rate, mutation rate, clonal rate and hypermutation rate. In this paper we investigate the effect of control parameters variation on the convergence speed for single and multiobjective optimization problems. Three examples are devoted for this purpose; namely the design of 2D recursive digital filter, minimization of simple function, and banana function. The effect of each parameter on the convergence speed of the IA is studied considering the other parameters with fixed values and taking the average of 100 times independent runs. Then, the concluded rules are applied on some examples introduced in [2] and [3]. Computational results show how to select the immune algorithm parameters to speedup the algorithm convergence and to obtain the optimal solution.

Despite the considerable amount of research related to immune algorithms and it applications in numerical optimization, digital filters design, and data mining, there is still little work related to issues as important as sensitivity analysis, [1][4]. Other aspects, such as convergence speed and parameters adaptation, have been practically disregarded in the current specialized literature [7][8]. The convergence speed of the immune algorithm heavily depends on its main control parameters: population size, replication rate, mutation rate, clonal rate and hypermutation rate. In this paper we investigate the effect of control parameters variation on the convergence speed for single and multiobjective optimization problems. Three examples are devoted for this purpose; namely the design of 2D recursive digital filter, minimization of simple function, and banana function. The effect of each parameter on the convergence speed of the IA is studied considering the other parameters with fixed values and taking the average of 100 times independent runs. Then, the concluded rules are applied on some examples introduced in [2] and [3]. Computational results show how to select the immune algorithm parameters to speedup the algorithm convergence and to obtain the optimal solution.

This paper presents one modern heuristic optimisation algorithm, named Taguchi-Based Immune Algorithm (TBIA), to solve the problem of designing 2D recursive digital filters with specified magnitude and group-delay characteristics. The algorithm is detailed for the design of three recursive filters’ categories, namely filters with predefined magnitude, delay and magnitude and delay. On the basis of minimising the magnitude and group-delay errors, multi-criterion design combination is employed to obtain optimal recursive filters that satisfy the required specifications. Computational experiments show the ability of the proposed algorithm to obtain more robust stable complex filters compared with previously reported design methods.

This paper presents one modern heuristic optimisation algorithm, named Taguchi-Based Immune Algorithm (TBIA), to solve the problem of designing 2D recursive digital filters with specified magnitude and group-delay characteristics. The algorithm is detailed for the design of three recursive filters’ categories, namely filters with predefined magnitude, delay and magnitude and delay. On the basis of minimising the magnitude and group-delay errors, multi-criterion design combination is employed to obtain optimal recursive filters that satisfy the required specifications. Computational experiments show the ability of the proposed algorithm to obtain more robust stable complex filters compared with previously reported design methods.