Aiming at the problems existing in the PID parameter tuning of traditional genetic algorithms, a method of applying adaptive genetic algorithms to parameter tuning was proposed. It takes system overshoot and dynamic performance indicators as the objective function, optimizes the crossover and selection probability in the genetic algorithm, reduces the probability of the system entering a local optimum, and makes the system converge faster. Comparing the traditional manual tuning PID and the genetic algorithm (GA) PID controller with the adaptive genetic algorithm (AGA) PID controller, it is concluded that the use of adaptive genetic algorithm can improve the performance indicators of the system.

No abstract

Flour plays an important role in People’s Daily consumption, and the content of impurities in flour indicates the quality of flour. At present, most domestic factories are using magnifying glass and other simple tools for impurity detection. This method is troublesome and does not meet the requirement of precision. This paper designs an automatic impurity detection system based on image processing, which not only improves the detection efficiency, but also greatly improves the detection accuracy. The basic process of this system is to grayscale the image obtained by photographing, then carry out local entropy transformation, and then map to form entropy image. Finally, the impurity detection is completed after image filtering, image segmentation and edge detection.

This paper presents two novel UHF RFID near-field reader antennas with uniform vertical electric field distribution. The two antennas have the following common characteristics. First, the radiating parts of the two antennas are simulated and fabricated by the microstrip lines and work using the leakage wave principle of microstrip lines. Second, the end of microstrip lines match the load to form a traveling wave mode of operation, so the two antennas have broadband characteristics. Third, both antennas are fed in a coaxial manner at the center of the antenna. The simulation and measurement results can show that the proposed three-branch antenna and four-branch antenna achieve good impedance matching in the range of 883–960 MHz and 870–960 MHz, respectively, and achieve uniform distribution of the vertical electric field component in a certain area. The reading areas of the three-branch antenna and the four-branch antenna are 70 mm × 70 mm × 90 mm and 100 mm × 100 mm × 120 mm (length × width × height), respectively. Due to the introduction of the ground plate, the antenna gain is low, which meets the design requirements of near-field antennas.