A Multi-CBR Algorithm Based on Comprehensive Evaluation for Operation Planning of Helicopter

Case-based reasoning (CBR) is an effective reasoning technology. The core idea of CBR is using past experience and expert knowledge to solve new problems. Operation planning of helicopter mainly relies on the experiences of the decision-maker, so CBR is an effective reasoning technology which can be used in the operation planning. Two key issues, evaluation method and case adaptation, are widely focused in CBR method. However, the existing CBR is not suitable for non-numeric optimization problem such as operation planning, because it belongs to small sample size problem. To solve this problem, the comprehensive evaluation based multi-CBR (CEB-mCBR) method is proposed in this paper, which can be used for operation planning of helicopter. According to the comprehensive evaluation theory, similar historical case set of the target case is established, and the historical operation plans are ranked. On this basis, the optimum initial plan for the target case is generated by the case screening CBR model. Then the solution element screening CBR model is further put forward to solve the issue of case adaptation. The CEB-mCBR method comprises multi-CBR (case screening CBR and solution element screening CBR) and combination evaluation method. It can integrate expert experience and human thinking better, especially when solution space is limited. Finally, through a case study analysis, the multi-CBR model can generate a feasible solution for the target case, and the solution elements can be adjusted according to the situation in the actual operation, which can improve the efficiency of operation planning while keeping the safety, economy and operating efficiency.