Research On The Parallel System Efficiency Evaluation And Application Based On Data Envelopment Analysis Method

In practical production activities,the production process of many decision-making units(DMUs)can be regarded as a production system with a network structure consisting of multiple subsystems.Among many types of network systems,parallel systems are a typical class of systems with similar subsystems,for example,the transportation system in China includes the passenger transportation subsystem and the freight transportation subsystem.The efficiency evaluation of parallel systems is important in economic production activities,not only to understand the current production situation,but also to provide suggestions for improvement of future production activities.Data Envelopment Analysis(DEA)is widely used to evaluate the efficiency of parallel systems because of its advantages such as no need to set production functions and prior weights.However,there are still some problems in the current research of parallel system efficiency evaluation based on DEA.Specifically,one is that the efficiency evaluation models for parallel systems under the traditional single frontier have some defects and do not fully consider the connection and difference between systems and subsystems.Another is that the efficiency evaluation of parallel systems does not take into account the heterogeneity of the technology environment in which the DMUs are located.Therefore,this paper takes the DMUs with parallel system as the research object,takes the existing DEA model as the basis,addresses the shortcomings of existing research,and conducts research on the efficiency evaluation of parallel system,which expands the theoretical research of DEA method and has certain guiding significance for the actual production decision-making of parallel system.The main research contents of this paper are as follows.(1)Research on the efficiency evaluation method of parallel systems under a single frontier.By reviewing the existing DEA models for parallel systems,the strengths and weaknesses of these models are summarized.Considering the problems that the existing models have low discriminatory ability for efficiency and can produce unreasonable inputoutput weights,as well as the problems that subsystems fail and unreasonable subsystem input-output weights are set without fully considering the connection and distinction between the system and subsystems,this paper proposes a neutral cross-efficiency evaluation method for parallel systems.The proposed method overcomes the shortcomings of the self-evaluation model and allows each subsystem to participate effectively in the evaluation by assigning different input-output weights to each subsystem in order to avoid subsystems being forced to operate under the same production mechanism.In addition,the proposed method is also applicable to special parallel systems with shared inputs.(2)Research on the efficiency evaluation method of parallel systems considering metafrontier.The current efficiency evaluation of parallel production systems does not take into account the heterogeneity of the technology environment and forms a reference benchmark that is difficult to achieve for many DMUs,which can lead to a waste of resources and is not conducive to the long-term development of DMUs.In view of the heterogeneity of the environment in which the DMUs are located,this paper extends the meta-frontier analysis to the parallel system efficiency evaluation method,proposes a group frontier efficiency evaluation model for DMUs with similar technology environments and a meta-frontier efficiency evaluation model for all DMUs,and decomposes the inefficiencies of the DMUs.The proposed method can improve the validity of the reference benchmark and can further determine the technology gaps in the system and subsystems of each DMU and provide directions for the technology improvement of DMUs.(3)Research on the evaluation of city innovation efficiency in China.There are two problems in the current research on city innovation efficiency,first,the problem of treating the innovation process of city innovation system simply as a black box without internal structure or using only the innovation process of knowledge or technology to represent the city innovation process;second,the problem of not taking into account the current situation of uneven development among cities in China.To address these problems,this paper divides the innovation process of China’s city innovation system into a parallel system composed of enterprise innovation subsystem,government and market innovation subsystem,and university and R&D institution innovation subsystem in parallel,proposes parallel DEA models for innovation efficiency evaluation of Chinese cities,summarizes the current situation and problems of city innovation system in China,and provides theoretical references and policy suggestions for the development of innovative cities in China.