Semua teknik pintar mempunyai kelebihan dan kelemahan. Oleh itu, teknik pintar sendiri tidak boleh digunakan universal kepada semua masalah sebab semua teknik pintar ada batasan terhadap masalah tertentu. Namum, beberapa teknik pintar telah digabungkan untuk menghasilkan satu hibrida sistem pintar. Gabungan dua atau lebih teknik pintar boleh mengatasi keterbatasan setiap teknik sendiri. Dalam projek ini, lapisan dikodekan riam optimasi (LECO) model telah dilaksanakan dalam masalah pengoptimuman reka litar. Evolusi mekanisme akan fokus pada Genetic Algorithm (GA) dan Particle Swarm Optimization (PSO). Gabungan GA dan PSO dalam berbeza struktur lapisan dalam LECO model telah dikaji. LECO model dilaksanakan dengan menggunakan MATLAB perisian. Optimasi pada litar reka dilaksanakan dengan menggunakan APLAC perisian. Prestasi LECO model dibandingkan dengan pengoptimasi dalam APLAC. Hasil maklumat dari LECO lebih baik daripada pengoptimasi kecerunan dalam mencapai tiga reka tujuan. Dari hasil maklumat, jelas bahawa LECO model boleh menangani masalah multi-keputusan dan multi-keberatan masalah pengoptimuman reka litar.
_________________________________________________________________________________________
Every intelligent technique has its own strengths and weakness. As such, individual intelligent techniques cannot be applied universally to solve all problems, since each of them has their own limitations towards certain problems. However, different types of intelligent techniques can be combined together to yield a hybrid intelligent system. Combination of two or more intelligent systems can overcome the limitations of each individual technique. In this project, Layer Encoded Cascade Optimization (LECO) model has been applied to optimization of circuit design problems. The evolutionary mechanism is focused on the Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) algorithm. Combination of GA and PSO in different layers of the LECO model is investigated. The LECO model is implemented using the MATLAB software. The optimization on circuit design is implemented using APLAC software. The performance of the LECO model is compared with those from built-in optimizers in APLAC. The results show that the LECO model is outperforms gradient optimizer in meet all the three design goals. From the results, it is evident that the LECO model is able to handle multi-decision and multi-objection optimization problems in circuit design.