工艺路线规划不仅受到加工设备方法，刀夹具选择的制约，也受到工艺设计原则

Process route planning not only by the processing equipment method, the tool holder restriction selection, but also by the process design principles (after the first rough finishing, after the first surface of the hole, once the first main reference advance, etc.) and the effects of a step of concentration and dispersion of [6, 7], in actual production, there may be a plurality of identical parts processing process plan feasible simultaneously. Therefore, under the premise of meeting the conditions of production and ensure product quality, technical and economic analysis and evaluation of different process scheme is to improve labor productivity, reduce processing costs, optimize resource utilization, an important way to reduce environmental pollution. At present, domestic and foreign scholars preferred aspects of the assessment process of planning a lot of research, it is generally agreed that a multi-constrained, non-linear combination of multi-objective optimization decision problems [8,9]. Traditional Machining process planning and program evaluation rely mainly on univariate knowledge and experience, not considering the interaction of various factors, leading to subjectivity and limitations applicable to the evaluation results. On this basis, scholars have put forward a variety of integrated assessment methods, such as multivariate regression analysis (MLR), artificial neural network, gray cluster analysis, genetic algorithm, ant colony algorithm, AHP, data envelopment analysis, etc. [10-14]. These methods typically employ separate programs appraised, achieved good results. However, due to mechanical product structure is becoming more complex, diverse mode of production affected by many factors, but with a lot of empirical evaluation, ambiguity and uncertainty, leading to the above method has its own advantages and limitations in the practical application. For example, multivariate statistical regression (MLR) can not solve the complex relationships among factors affecting highly nonlinear; initial threshold artificial neural network is difficult to determine the number of samples need to be trained in the fitting process and is easy to fall into local optimal solution, resulting in model the lack of generalization; the result of low-resolution gray clustering method, sometimes inconsistent with the actual situation; indexes by the excessive weight of large and statistics is difficult to determine, experts give different index score of causing a variety of different evaluation As a result, the actual production to increase even more difficult. Fuzzy comprehensive evaluation fuzzy transformation theory and membership theory, can fully consider the mutual coupling between the influencing factor, effectively solve the fuzzy impact assessment criteria border, so that the evaluation complete and accurate, widely used in engineering technology [ 15, 16]. The method weights the weight coefficient, and the comprehensive evaluation membership function selection mode directly affects the accuracy of the evaluation. On the basis of comprehensive consideration of various factors technical feasibility, economy, efficiency and environmental effectiveness of the correction traditional fuzzy comprehensive evaluation of the weight assignment and fuzzy operation mode, the establishment of an improved fuzzy comprehensive evaluation and the level of entropy multi-process planning scheme evaluation model analysis combining with process planning and car adapter frame parts as an example evaluation and analysis in order to provide guidance for the comprehensive evaluation of preferred machining process planning programs in order to achieve shorter product cycle times, improve resource optimize the use, production costs and reduce the efficacy.

Process route planning is not only subject to the processing equipment method, knife fixture selection constraints, but also by the principles of process design (first coarse, then fine, first face after hole, first master, second, benchmark first, etc.) as well as the impact of process concentration and dispersion, in the actual production, the same part processing may exist at the same time a number of feasible process planning programs. Therefore, under the premise of meeting the production conditions and ensuring the quality of production of enterprises, it is an important way to improve labor productivity, reduce processing costs, optimize resource utilization and reduce environmental pollution. At present, scholars at home and abroad have done a lot of research on the evaluation and optimization of the process planning scheme, which is generally considered to be a multi-binding, nonlinear, multi-objective combination optimization decision-making problem. The traditional machine-plus process planning program evaluation mainly relies on single factors and empirical knowledge, without comprehensive consideration of the combined effect of various influencing factors, resulting in the subjective and applicable limitations of evaluation results. On this basis, scholars put forward a variety of comprehensive evaluation methods, such as multi-regression analysis (MLR), artificial neural network method, gray clustering method, genetic algorithm, ant colony algorithm, hierarchical analysis method, packet analysis method, etc. These methods are usually used alone to evaluate the scheme, and good results are obtained. However, due to the increasingcomplexit of mechanical product structure, the variety of production methods, many factors, and many evaluation indicators with experience, ambiguity and uncertainty, resulting in the above methods in practice have their own advantages and limitations. For example, the multi-statistical regression method (MLR) cannot solve the complex relationship of high nonlinearity between the factors, the initial threshold of the artificial neural network is difficult to determine, the number of samples needed to be trained is large and the local optimal solution is easily trapped in the fitting process, resulting in insufficient generalization ability of the model, and the low result resolution of the gray clustering method. Sometimes does not agree with the actual situation, when there are too many indicators of hierarchical analysis, the statistics are large and the weight is difficult to determine, and the different index scores given by different experts result in a variety of evaluation results, and even increase the difficulty of actual production. Based on the theory of fuzzy transformation and membership, fuzzy comprehensive evaluation method can give full consideration to the mutual coupling of the influencing factors, effectively solve the influence of the fuzzy boundary of the evaluation criteria, make the evaluation results comprehensive and accurate, and be widely used in engineering technology. The selection of weight coefficient, membership function and comprehensive evaluation method in this method directly affects the accuracy of evaluation. Based on the comprehensive consideration of various influencing factors such as technical feasibility, economy, efficiency and environmental benefits, this paper revises the weight assignment and fuzzy operation method of the traditional fuzzy comprehensive evaluation index, and establishes a multi-process planning program evaluation model to improve the combination of fuzzy comprehensive evaluation and hierarchical entropy analysis. The process planning of the automotive adapter parts is used as an example for evaluation and analysis, with a view to providing guidance for the comprehensive evaluation and selection of the machine-plus process planning scheme, in order to shorten the product processing cycle, improve the optimal utilization of resources, and reduce the efficiency of production costs.

The process planning is not only restricted by the selection of processing equipment and tool fixture, but also influenced by the process design principles (first coarse, then fine, first face, then hole, first primary, then secondary, first benchmark, etc.) and the concentration and dispersion of processes [6,7]. In actual production, there may be multiple feasible process planning schemes for the same part processing at the same time. Therefore, on the premise of satisfying the production conditions and ensuring the production quality of the products, the technical and economic analysis and comprehensive evaluation of different process schemes is an important way to improve labor productivity, reduce processing costs, optimize resource utilization and reduce environmental pollution. At present, scholars at home and abroad have conducted a lot of research on the evaluation and optimization of process planning scheme, which is generally considered as a multi constraint, non-linear, multi-objective combination optimization decision-making problem [8,9]. The traditional evaluation of machining process planning scheme mainly relies on single factor and empirical knowledge, and does not consider all kinds of influencing factors together, which leads to the subjectivity and applicability limitations of the evaluation results. On this basis, scholars put forward a variety of comprehensive evaluation methods, such as multiple regression analysis (MLR), artificial neural network, grey clustering, genetic algorithm, ant colony algorithm, AHP, DEA and so on [10-14]. Generally, these methods are used alone to evaluate the scheme, and good results are achieved. However, due to the increasingly complex structure of mechanical products, various production methods, many influencing factors, and many evaluation indexes with experience, fuzziness and uncertainty, the above methods have their own advantages and limitations in practical application. For example, multivariate statistical regression (MLR) can not solve the highly nonlinear complex relationship between influencing factors; the initial threshold of artificial neural network is difficult to determine, the number of samples to be trained is large, and it is easy to fall into local optimal solution in the fitting process, resulting in insufficient generalization ability of the model; the result resolution of grey clustering method is low, sometimes inconsistent with the actual situation; analytic hierarchy process When there are too many indicators, the data statistics are large and the weight is difficult to determine. Different experts give different index scores, resulting in a variety of evaluation results, even increasing the actual production difficulty. Based on the fuzzy transformation principle and membership degree theory, the fuzzy comprehensive evaluation method can fully consider the interaction between the influencing factors, effectively solve the influence of fuzzy evaluation standard boundary, make the evaluation results comprehensive and accurate, and widely used in the field of engineering technology [15,16]. In this method, the selection of weight coefficient, membership function and comprehensive evaluation method directly affects the evaluation accuracy. Based on the comprehensive consideration of various factors such as technical feasibility, economy, work efficiency and environmental benefit, this paper revises the weight assignment and fuzzy operation mode of traditional fuzzy comprehensive evaluation indexes, establishes a multi process planning evaluation model combining improved fuzzy comprehensive evaluation and hierarchical entropy weight analysis, and evaluates the process planning of automobile transfer frame parts as an example Price analysis is expected to provide guidance for the comprehensive evaluation and optimization of machining process planning scheme, so as to shorten the product processing cycle, improve the optimal utilization of resources, and reduce the production cost.<br>