Earticle

To solve the "soft field" effect and the ill-posed problem in electrical capacitance tomography technology, a novel image reconstruction algorithm based on population entropy and adaptive differential evolution for Electrical Capacitance Tomography is proposed in this study. The algorithm uses all the gray pixels as the initial population’s individual. After finite iterations, the algorithm mutates and makes crossover of the population in order to obtain the optimal species populations. That is the optimal value for the ECT imaging pixels. The population entropy and the variation factor make the range of each searching generation decreasing. In the simulation, the improved adaptive differential evolution algorithm will be compared with the LBP algorithm. The result shows that the new algorithm has better image quality and more stable boundary than the LBP Algorithm, which provides a new way to reconstruct images for ECT.

To solve the slow convergence speed, low precision in later period and tedious parameter setting of differential evolution when applied to complex optimization functions, an improved differential evolution algorithm (dn-DADE) based on dynamic adaptive strategy is proposed. Firstly, the elite solutions of current population are utilized in the new mutation strategy (DE/current-to-dnbest/1) to guide the search direction, and then these optional elite solutions tend to the global optimal solution in the late stage of evolution to balance the diversity of population and convergence speed. Secondly, the adaptive update strategies of scaling factor and crossover factor are designed for control parameter values self-adapting at different search stages, thus improve the stability and robustness of the algorithm. A set of 14 benchmark functions is adopted to test the performance of the proposed algorithm. The results show that dn-DADE algorithm has the advantages of remarkable optimizing ability, higher search precision, faster convergence speed and outperforms several state-of-the-art improved differential evolution algorithms in terms of the main performance indexes.

이 연구는 단봉 전역 최적화 성능이 개선된 적응적 코시 분포 차분 진화 알고리즘을 제안한다. 기존 적응적 코시 분포 차분 진화 알고리즘은(ACDE) 개체의 다양성을 보장하여 다봉 전역 최적화 문제에 우수한 "DE/rand/1" 돌연변이 전략을 사용했다. 그러나 이 돌연변이 전략은 수렴 속도가 느려 단봉 전역 최적화 문제에 단점이 있다. 제안 알고리즘은 "DE/rand/1" 돌연변이 전략 대신 수렴 속도가 빠른 "DE/current-to-best/1" 돌연변이 전략을 사용했다. 이때, 개체의 다양성이 부족하여 발생할 수 있는 지역 최적해로의 수렴을 방지하기 위해서 매개변수 초기화 연산이 추가됐다. 매개변수 초기화 연산은 특정세대를 주기로 실행되거나 또는 선택 연산에서 모든 개체가 진화에 실패하는 경우 실행된다. 매개변수 초기화 연산은 각 개체들의 매개변수에 탐험적 특성이 높은 값을 할당하여 넓은 공간을 탐색할 수 있도록 보장한다. 성능 평가 결과, 개선된 적응적 코시 분포 차분 진화 알고리즘이 최신 차분 진화 알고리즘들에 비해 특히, 단봉 전역 최적화 문제에서 성능이 개선됨을 확인했다.

In this paper, an improved ACDE (Adaptive Cauchy Differential Evolution) algorithm with faster convergence speed, called ACDE2, is suggested. The baseline ACDE algorithm uses a "DE/rand/1" mutation strategy to provide good population diversity, and it is appropriate for solving multimodal optimization problems. However, the convergence speed of the mutation strategy is slow, and it is therefore not suitable for solving unimodal optimization problems. The ACDE2 algorithm uses a "DE/current-to-best/1" mutation strategy in order to provide a fast convergence speed, where a control parameter initialization operator is used to avoid converging to local optimization. The operator is executed after every predefined number of generations or when every individual fails to evolve, which assigns a value with a high level of exploration property to the control parameter of each individual, providing additional population diversity. Our experimental results show that the ACDE2 algorithm performs better than some state-of-the-art DE algorithms, particularly in unimodal optimization problems.