Abstract: Elevators are a common vertical transportation tool in high-rise buildings. To improve the operating efficiency and service quality of multiple elevators, adopting a high-quality elevator group control system to uniformly manage the operation of multiple elevators is a worthwhile research problem. This paper introduces a control algorithm for elevator group control systems, namely the genetic algorithm. Taking the scheduling of four elevators as a background, a corresponding fitness function is established, and the genetic algorithm is used to implement the elevator scheduling scheme. Simulation experiments show that this elevator scheduling method is effective. Keywords : Elevator group control system; Genetic algorithm; Fitness function [align=center] Design of Remote Elevator Monitory System ZHANG Jin-yang Chen Fei (Xinjiang Uygur Autonomous Region special equipment examination research institute, 830000) order to raise many elevator operating efficiency and the grade of service, uses the high quality elevator group control system, the global administration many elevators movement is the question which is worth studying. This article introduced in the elevator group control system's one kind of control algorithm is the genetic algorithm, through four elevator's dispatches is the background, establishes the corresponding sufficiency function, and uses the genetic algorithm to realize the elevator dispatch plan. The simulation experiment indicated that this elevator dispatch method is effective. Key words : elevator group control system; genetic algorithm; sufficiency function 1 Introduction Elevators, as a vertical transportation tool, are now widely used. With the increasing number of high-rise buildings, these structures often require several or even dozens of elevators to meet passenger needs. While increasing the number of elevators can improve operational efficiency to some extent, the key to efficient elevator operation lies in the ability to manage passenger flow. Since the invention of the elevator, from single-elevator operation to parallel operation of two elevators, and then to elevator group control systems, a traffic configuration theory for elevators has gradually formed. As the demand for elevators continues to grow, the analysis, design, and scheduling algorithms of elevator group control systems have become increasingly prominent. Furthermore, due to the inherent randomness and nonlinearity of the system, and the diversity of control objectives, the scheduling system has become extremely large, and the scheduling algorithms have become increasingly complex. This necessitates the effective improvement and development of elevator group control systems using intelligent control technology. The scheduling method is the core of the elevator group control system, directly affecting the operation of each elevator and the quality of service provided by the elevator system. With the rapid development of artificial intelligence theory, various intelligent elevator scheduling methods have emerged, such as elevator scheduling methods based on fuzzy models, expert-based elevator scheduling methods, neural network-based elevator scheduling methods, and genetic algorithm-based elevator scheduling methods. Currently, most elevator group control scheduling problems in China are solved using fuzzy neural network technology, while the application of genetic algorithms to elevator group control theory is a current research hotspot. Because genetic algorithms are independent of gradient information when searching for optimal solutions, automatically acquire and accumulate knowledge about the search space during the search process, adaptively control the search process, and are simple, universal, robust, and suitable for parallel distributed processing, they have great application prospects in elevator group control theory. 2. Operation of Genetic Algorithm in Elevator Group Control System This algorithm transforms the parameters or solutions of the search space during the objective optimization process into chromosomes in the genetic space. A certain number of chromosomes constitute the initial population. A fitness function is constructed based on the objective optimization function, and the fitness function value of each chromosome is calculated. Then, chromosomes are selected based on the fitness function values, and crossover and mutation operations are performed with a certain probability to generate new chromosomes, forming the next generation of the population. This process continues until the optimal solution is found or a sufficient number of generations have been reached. Considering the real-time nature of the elevator group control system, each time the genetic algorithm is called for a search, only a few searches are performed within a finite time, rather than obtaining a convergence value every time. Although this doesn't always yield the optimal value, considering the randomness of the elevator group control system, the optimal value isn't particularly meaningful. New floor call signals are generated constantly, and other external conditions may change at any time. Even if the optimal allocation scheme is found at the current moment, it may no longer be optimal under new conditions. When no new floor call signals are generated, the elevator group control system periodically calls the genetic algorithm again to search based on the current system state, assigning service elevators to all unresponsive floor call signals. The overall flowchart of the genetic algorithm operation is shown in Figure 1. [align=center] Figure 1 Overall Flowchart of Genetic Algorithm Operation[/align] In this design, chromosomes use integer binary encoding, encoding the elevator numbers. There are 4 elevators, so elevators 1-4 are encoded as 00, 01, 10, and 11, respectively. Each unassigned call signal in the elevator system corresponds to a 2-bit binary number, indicating that the elevator corresponding to the encoded value will respond to the call signal. The chromosome length is twice the number of currently unresponsive floor call signals, i.e., a variable-length chromosome is used, with the length varying with the number of floor call signals. This approach has two advantages: first, it avoids using longer chromosomes, reducing computational complexity; second, it prevents invalid solutions. During each optimization, if there are M unresponsive floor call signals, the chromosome is represented by an integer code string of length 2m. A chromosome represents a dispatch scheme of the group control system for the current outgoing call signal. Taking four 15-floor elevators as an example, an array C[0...27] records the assigned floor call signal numbers, with calls to floors 1-14 recorded as 0-13 and calls to floors 2-15 recorded as 14-27. If the values ​​of C[0] to C[5] are 1, 6, 12, 18, 19, 26 respectively, specifically indicating calls to floors 2, 7, and 13 to floors 6, 7, and 14 to floors 14, this corresponds to a chromosome encoded as 341232. 3. Fitness Function Design Genetic algorithms generally do not require other external information during the search and evolution process. They only use the evaluation function value to evaluate the quality of individuals or solutions, which serves as the basis for subsequent genetic operations. It is important to note that the evaluation function and the fitness function are not the same concept. The evaluation function refers to the objective function of the optimization problem; the optimization process using a genetic algorithm is essentially finding the extreme value of the evaluation function. The fitness function, on the other hand, is a function mapped from the evaluation function to facilitate comparison of individual sizes and selection, crossover, and mutation operations. The fitness function value is called fitness, which indicates the strength of an individual's adaptability to the environment. In evolutionary search, genetic algorithms rely on the size of the fitness function value to distinguish the quality of each individual; individuals with higher fitness values ​​have a greater chance of reproducing. The fitness function evaluation is the basis for selection operations. In practical applications, the design of the fitness function must be combined with the requirements of the problem being solved. Generally, it can be derived from the objective function of the problem. This paper takes the group control of 4 elevators as the research object, and constructs the fitness function: (3) where a determines the mandatory selection. The smaller a is, the greater the difference between the new fitness of the individual with higher original fitness and the new fitness of other individuals, which increases the mandatory selection of that individual. 4. Simulation of elevator dispatching based on genetic algorithm To test the performance of the group control algorithm, MATLAB is used to implement the virtual simulation of elevator dispatching based on genetic algorithm. The elevator configuration parameters of the group control system are set as follows in the simulation experiment: the group control system consists of 4 elevators with 15 floors, a speed of 2 m/s, an acceleration of 1.5 m/s², a door opening time of 1.5 seconds, a door closing time of 3 seconds, and a building floor height of 3 meters. In order to test the dispatching algorithm, a heavy traffic flow between floors is randomly generated. The following simulation is carried out under this system parameter and traffic flow. [align=center]Figure 2. Elevator operation curve based on genetic algorithm dispatching[/align] As can be seen from Figure 2, in most cases, the distribution of elevators across floors is relatively uniform, with no clustering phenomenon. The number of elevators going up and down is relatively even, which is a reasonable traffic pattern. 5. Conclusion This paper proposes a scheduling method for elevator group control systems based on genetic algorithms. This method can optimize the dispatching of multiple elevators to improve the overall service performance of the system and thus obtain better dispatching results. References 1. Zhou Ming, Sun Shudong. Principles and Applications of Genetic Algorithms [M]. Beijing: National Defense Industry Press, 2003. 2. Atsuya Fujino, Toshimitsu Tobita, et al. An elevator group control system with floor-attribute control method and system optimization using genetic algorithms [J]. IEEE Trans. On Industrial Electronics, 1997, 44(4): 1502-1507. 3. Wang Xiaoping, Cao Liming. Genetic Algorithms [M]. Xi'an: Xi'an Jiaotong University Press, 2005. Contact number: 0991-5853331 ext. 8308 Address: No. 9, Mianhua Street, Changjiang Road, Urumqi Unit: Xinjiang Special Equipment Inspection and Research Institute