Evaluation on Indicator System of Food Economic Project Evaluation based on Combination Weight and TOPSIS

This study takes the drawback that the project performance has great influence on food enterprise performance into consideration, through the introduction of financial indicators, indicators of internal processes, learning and development indicators, to build a new comprehensive performance evaluation index system for food economic projects. Based on this, performance indicators data of food economic project is collected and then further processed, which lays the foundation for determining indicators weights.


INTRODUCTION
The evaluation of Food economic Project is adaptation of socialism food economic market development, is consolidation of investing and financing reform and is meaningful to the realization of scientific project decision-making, enhancing the overall benefit of the project and implementing the strategic target of sustainable development of our country (Herman, 2008;Fu and Xu, 2008). With the development of investment system reform and the establishment of restrictive mechanism venture, the input-output of investment project and the successful rate of investment decision-making will be further emphasized in a variety of investors, the dissertation makes a detailed and systematic study of the theory and method of food economic project evaluation, which aims to services supervision and evaluation of our domestic Food economic Project with elements in both theory and technology and the system is valuable in practice (Zavadskasn et al., 2014). Traditional indicators of performance evaluation system involves mostly just in quality, schedule and environmental goals of the project, that is to say, it only considers the level of performance that only a few large projects in progress goals, without taking into account of the impact on the overall performance level caused by the staff involved in the project and project management level of internal management of the project, which cannot meet the requirements for the performance evaluation (Wang and Lai, 2008).

Processing of indicators data:
Unification method: According to the indicator system established in this study, the indicators including income satisfaction of staff, capital utilization, administration innovation, management quality, staffs' learning attitude, perfect degree of incentive mechanism, timeliness of material, equipment and food economic units, completion of food economic schedule plan, complete food economic data, reduction rate of project cost, contract compliance rate, resolution rate of technical problem, passing rate of project quality, the quality qualification rate of raw materials and equipment, environmental assessment level are benefit indicators; budget and final accounts deviation rate, environmental protection input cost indicators (Ngacho and Das, 2014).
Cost indicators are transferred into benefit indicators in this study (Sun and Luo, 2008). The transfer formula is shown as: Non-dimensional method: When evaluating the project performance level, different dimensions of different indicators makes it hard for the comprehensive evaluation (Wen and Ren, 2011). Then, nondimensional method is applied to unify the indicators dimensions, where efficacy coefficient is introduced in this study (Lv, 2009). The formula of efficacy coefficient is shown below: where, Mj, mj respectively represent maximum and minimum values of each indicator; c and d are known constants, which make the numbers for transformation translate and zoom. The c and d are respectively valued 60 and 40 in this study.
Eigenvalue method: Eigenvalue method is used for determining indicators weights by establishing evaluation matrixes and comparing evaluation indicators. Firstly, evaluation matrix A is established and the eigenvector of matrix A can be calculated by the following formula (A-λE)x = 0, where λ is the root of equation |A-λE| = 0 and then the indicators weights sequencing can be obtained. The eigenvector corresponding to the maximum eigenvalue of matrix A is the weight vector wanted. However, this method needs examining consistency. Consistency indicator CI can be calculated with the following formula CI= (λ max -m)/(m-1), where m represents the matrix order. The average random consistency RI meets demands of the Table 1.
The random consistency rate CR can be calculated with formula CR = CI/RI. When CR is lower than 0.10, the judgment matrix has satisfied consistency; or the judgment matrix need adjusting, until it has satisfied consistency.
Entropy weight method: Entropy weight method is used to determine the indicators weights based on information load of each indicator (Qu and Fang, 2014). Here establishes matrix R = (x ij ) m*n , in which x ij represents the ith indicator attribute corresponding to jth item. The detailed procedure of the entropy weight method is given below: where, p ij represents the contribution degree of the jth item corresponding to ith indicator attribute.

Entropy:
Entropy e i is the total contribution of all the items to ith indicator: Where, constant k can be calculated by formula k = 1/Inn. Difference coefficient: g i represents contribution difference of each item corresponding to the ith indicator: Indicators weights: The entropy weight of the jth indicator can be obtained by the following formula: Combination weight method: In order to avoid the influence of experts' subjective preferences in subjective weight and uncontrolled weighting result in objective weight, the combination weight is applied to determine the indicators weights (Song and Yang, 2004). The weight vector by eigenvalue method is shown as  and the weight vector by entropy weight method is shown as Then the combination weight of each indicator can be calculated with the following formula: where, k 1 + k 2 = 1 and k 1 , k 2 > 0.
TOPSIS: TOPSIS is originated in discrimination problems of multivariate statistical analysis (Luo and Peng, 2011). The detailed procedures of TOPSIS are shown below: Standard decision matrix: Based on the indicators data after processing in Table 1 to 5, the standard decision matrix T = (t ij ) m*n can be established.
Weighted decision matrix: Based on the indicators combination weight vector W = (w1, w2, …, wm), the weighted decision matrix can be obtained by the following formula: where, w j is the weight of the jth indicator and 1 1 n j j w    .
Positive and negative ideal points: In the weighted decision matrix, the positive ideal point x + is a vector composed of maximum number of each column; the

Distances with positive and negative ideal solutions:
The n-dimension Euclidean distances between evaluated object and positive, negative ideal points are shown by and , which are calculated with the following formulas: It is obvious that the larger C i is, the longer distance between evaluated object and negative ideal solution and shorter distance between evaluated object and positive ideal solutions are. Then, rank the similarities of all the objects and select the best evaluated object.

RESULTS AND DISCUSSION
Indicators data processing: When conducting unification of indicators data, M 51 and M 81 are respectively valued 1.40 million and 100%. Then based on the unification result, the efficacy coefficient method is applied to unify the indicators dimensions. The processing result is shown in Table 2.

Determining the indicators weights:
Indicators weights by eigenvalue method: According to the eigenvalue method, the indicators scores tables of             Table 3 to 11. And the weight of each indicator relative to the first class indicator is calculated, as Table  12 shows. The maximum eigenvalue λ max of judgment matrix is 8.2641, the value of Consistency indicator CI0.0377, the value of random consistencyRI0.0267 and the matrix has satisfied consistency.

Indicators weights by entropy weight method:
Based on the indicator system and indicator data from three projects, the entropy weight method is applied to determine indicators weights, as Table 13 shows.

Indicators combination weights:
The weights of subjective and objective weight methods for the combination weight are calculated by lagrange algorithm and the result shown in Table 14.
The result of TOPSIS: Weighted judgment matrix can be calculated by the indicators data after processing and indicators combination weights, as the Table 15 shows.
According to the Table 16, the score of P2 is the highest. And by the methods applied in this study, the performance levels of three projects are evaluated, in which P2 ranks first, P3 second, P1 third.

CONCLUSION
In this study, the project and construction project is described, the relation between project evaluation and project life-cycle and the significance of making evaluation are analyzed, developing process of evaluation is reviewed; On the basis of analysis, generalization and summarization, the theory basis of construction project food economic evaluation, environment evaluation and social evaluation are put forward to make a basis on further study; the concept, basic process, the principle and the basic content of the comprehensive evaluation of construction project are introduced, then the establishment and standard way of index system in the Multiple Attribute Evaluation (MAE) of construction project are studied; Based on the analysis of subjective weights and objective weights, the subjective and objective synthetic method, which makes up the default of subjective weights or objective weights, is put forward; Through comprehensive and thorough research, the classification of the commonly used evaluation method is carried on and the choice principle of evaluation method is elaborated, which provides the basis for the people to select it correctly; According to the research results of index system of construction project comprehensive evaluation, on the foundation of the introduction of construction evaluation content, the food economic evaluation, the society and the environmental effect evaluation index system are established; In profits from the most recent information related of the domestic and foreign on project evaluation and in the reference of various research results related food enterprise achievements evaluation, a set of index system to evaluate the food enterprise. At last, TOPSIS is applied to comprehensively evaluate the projects performance levels.