Abstract
The search range and search position of the particle are closely related to the shrinkage-expansion factor of the particle and the value of the center of the potential well. As the number of iterations increases, its search will gradually fall near the center of the potential well, and the search range will gradually decrease. Therefore, when the center of the potential well and the search range gradually approach the global optimum, the final result of the algorithm can be guaranteed to be the global optimum. Therefore, we need to optimize the shrinkage-expansion factor and the potential well center, improve the individual development ability of the group in the later stage of the algorithm and the expansion ability of the group, so that the potential well center is easier to fall near the global optimal point. Based on the chaotic strategy, a particle swarm initialization method is proposed. Using the unique ergodicity of the chaotic system, the method can make the particles spread well in the solution space of the search problem. Based on the antecedent normal cloud model, an adaptive method for determining control parameters and the center of the potential well is proposed. Based on the consequent normal cloud model, a particle adaptive mutation method is proposed. The comprehensive application of these improvement measures to the improved QPSO effectively improves the performance of the original algorithm. In this paper, the principle and method of multioptical axis parallelism measurement are studied. Based on the hardware equipment of the parallelism measurement system based on the large-diameter off-axis parabolic mirror collimator, the measurement accuracy and automation of the system are improved through image processing technology. The degree and anti-interference ability of the system are analyzed in detail. According to the characteristics of the system, an optimized focus evaluation function and a center positioning algorithm are proposed to improve the measurement speed and accuracy of the system. Through the improvement of the system DCT focus evaluation coefficient method and the optimization of the least squares ellipse fitting process, the measurement accuracy and anti-interference ability of the system are improved, the calculation time is shortened, and the real-time performance and automation of the system are enhanced.
Highlights
In order to further prevent the convergence speed of particles from slowing down in the evolution process, especially in the case of falling into a local optimum and the convergence accuracy cannot continue to be optimized due to insufficient diversity of particles in the later stage of the algorithm, this paper introduces a normal cloud model
It is possible to improve the late-stage development capability of the algorithm and increase the activity of particles by improving the center of the potential well and the average best position, from the perspective of the operation process of the QPSO algorithm, the particle is only centered on the attractor, and the average best position is the radius
In the later stage of the algorithm, there is no mechanism to mutate the particle itself, increasing the ability of individual particle development that gradually weakens with iteration. is ability is an important factor for the algorithm to jump out of the local optimum. erefore, it is necessary to find a mutation method, which increases the search ability of particles in the later stage of the algorithm, so that the algorithm is not easy to fall into the local optimum. ere is an important problem in the QPSO algorithm and the PSO algorithm, that is, there will be premature phenomenon in the search process of the particles
Summary
Computing is one of the most important aspects of human thinking ability, and the improvement of computing ability is closely related to the progress of human civilization [1]. E modern information processing with computer as the core brings human beings into a new information age. Erefore, accelerating the computing speed of the computer to improve the computing power of the computer has become one of the central tasks of computer science [2]. Is problem can be solved from two aspects. One is to manufacture more advanced computer hardware, and the other is to design appropriate computer operation processes, which can be called “algorithms.” e proposal of intelligent computing provides us with new ideas for finding fast algorithms and solving complex problems. Because computational intelligence does not need to establish an accurate mathematical model of the problem itself, it is suitable for solving those
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
