Abstract: Aiming at the performance problems of new R&D institutions, such as the uncertainty of performance evaluation value, the correlation between indicators, the diversification of investors with risk preference, and the unknown weight of indicator set, a performance evaluation method based on double reference points and Choquet integral is proposed. Firstly, combined with the characteristics of new R&D institutions, the performance evaluation index system is constructed; secondly, the double reference points of peers and expectations are designed based on the cumulative prospect theory; then, the direct correlation matrix between the indexes is constructed, and the weight of the index set is optimized according to the k-additive fuzzy measure and the Banzhaf value of the average marginal contribution, which is substituted into the Choquet integral method and the reference point weight formula to calculate the comprehensive performance evaluation value of each new R&D institution, so as to realize the organization ranking and problem analysis. The method considers the characteristics of new R&D institutions, data uncertainty, index relevance, risk preference, weight unknown and other factors, which is more in line with the actual situation. The case verifies the effectiveness and practicability of the method.

Key words: new R&D institutions, performance evaluation, double reference points, Choquet integral, the weight of index set

摘要： 针对新型研发机构绩效评估值不确定、指标间存在关联、投资主体多元且有风险偏好、指标集权重未知等绩效问题,提出基于双参照点和Choquet积分的绩效评估方法。首先,结合新型研发机构的特点,构建绩效评估指标体系;其次,基于累积前景理论设计同行-期望双参照点;然后,构建指标间直接关联矩阵,依据K-可加模糊测度和平均边际贡献Banzhaf值优化求解指标集权重,代入Choquet积分方法和参照点权重公式计算各新型研发机构的综合绩效评估值,从而实现机构排序和问题分析。该方法考虑了新型研发机构的特点、数据不确定性、指标关联性、风险偏好性、权重未知性等因素,更加符合实际情况,案例验证了方法的有效性和实用性。

关键词: 新型研发机构, 绩效评估, 双参照点, Choquet积分, 指标集权重