关键词: 维修战斗力, 资源配置, 混合整数, 非线性规划, EGA, ε约束准则

Abstract: Ship maintenance resource allocation problem is essentially the extension and application of resource scheduling problem in ship maintenance field. Compared with other military equipment, ship equipment is a typically sophisticated large complex weapon equipment with large unit price, high scientific and technological integration, as well as complex technology. Therefore, it has higher internal requirements for optimizing ship maintenance resource allocation. Therefore, overall consideration of multi-dimensional resource influences and research on optimal design of ship maintenance resource allocation scheme have high practical application value for improving the efficiency of ship maintenance task scheduling, enhancing the economic benefits of ship repair units and maximizing the combat effectiveness of repaired ships.
Against the background of multi-dimensional resource integration in ship maintenance, in order to improve the scientific and effective allocation of ship maintenance resources, this paper proposes the definition of maintenance combat effectiveness in view of ship maintenance objectives, and regards the generation of maintenance combat effectiveness as the combination of the quantity and quality of various maintenance resources. Considering the influence of financial, human, material and technical resources on the allocation of ship maintenance resources, a ship maintenance resource allocation model is constructed. In the modeling design, taking the maximum military benefit (maintenance combat effectiveness) and the minimum maintenance cost as the optimization objectives, the model parameters and decision variables are defined scientifically, and the ship maintenance resource allocation problem is transformed into a multi-objective optimization problem of mixed integer nonlinear programming model. In terms of solving ideas and methods, based on the standard genetic algorithm(SGA), ε-constraint criterion of multi-objective decision is introduced to ensure the non-uniquenness of model solution results under different constraints, so as to form a decision scheme set for decision analysis. At the same time, in order to further improve the convergence speed and stability of model optimization, the elite retention strategy is introduced into the basic algorithm, and a new ε-EGA multi-criterion adjustment algorithm is constructed. The solving steps of the complete elite retention genetic algorithm under the ε-constraint criterion are provided to ensure that satisfactory Pareto solution sets and frontiers can be searched more quickly.
Finally, an empirical test is carried out based on the actual annual ship maintenance tasks of enterprise H, so as to improve the scientific nature of the conclusions of the optimal allocation model and enhance the consistency between the model and the actual situation. The data involved in the demonstration, such as ship maintenance funds, personnel type and quantity, maintenance equipment type and quantity, are provided by the financial management department of Enterprise H. The analysis results show that when the ε-EGA multi-criterion adjustment algorithm is running, the fitness function value of the optimal individual decreases with the increase of population genetic algebra, and is basically stable when the population evolves to 120 generations. The optimal fitness value of the population after the evolution to 200 generations is -347.87, slightly less than the average fitness value -345.09. It indicates that the overall evolution level of the population is high and the optimized decision scheme can reach the predetermined goal. The results show that the ε-EGA multi-criteria adjustment algorithm has good applicability, fast calculation speed and high optimization degree, and plays an obvious role in the allocation of ship maintenance resources of enterprise H. ε-EGA multi-criteria adjustment algorithm can provide dynamically changing optimal decision scheme synchronously according to the adjustment and change of constraint criteria, and the method has strong stability. The maintenance resource allocation model established in this paper also means significant for resource scheduling and planning in other engineering industries.
Since the influence of maintenance time on resource allocation is not considered in the process of resource allocation optimization, the proposed decision-making method of ship maintenance resource allocation based on &-EGA multi-criteria adjustment still leaves room for further optimization and discussion. If the maintenance time variable is included in the model design as an influencing factor, the adaptability of the model can be further improved from the perspective of maintenance resource use sequencing. This paper has been greatly supported by the Youth Project of the National Social Science Foundation(19CGL073)and the Department of Management Engineering and Equipment Economics of Naval University of Engineering. We would like to express our deep gratitude here.

Key words: maintenance combat effectiveness, resource allocation, mixed integer, nonlinear programming, EGA, ε-constraint criterion