Abstract: Post-crisis era, the trust and reputation problems in the cooperative enterprises have highlighted the risk of virtual enterprise, Virtual Enterprise trust evaluation model is presented based on QFD method, trust evaluation set and trust evaluation grades at first. Then, project schedule optimal model base on reputation is given integrating push and pull inventory mechanism which reflects that reputation is the center of modern management concepts according to the principle of maximizing long-term interests. Finally, the relative optimal algorithm is discussed for implement during partner selection and profits distribution of virtual enterprise. Furthermore, decision-making model is given by Reputation-based incentive mechanisms.
INTRODUCTION
During the second half of 20th century, the rapid development of information technology promotes the globalization of markets. The users have diversified and personalized demands so that the product life cycle has become shorter and shorter. In order to adapt to the flexible market, virtual enterprise emerged and has been considered to be the main form of enterprise organization in 21th century (Bishr and Mantelas, 2008). Based on the mutual trust, it establishes a long-term alliance whose members include manufacturers, suppliers, distributors and customers (Porter, 1990). “Mutual trust” is not only the basis of cooperation and mutual success but also the precondition of the realization of the agility of virtual enterprise. The “long-term alliance” is abound to accumulate the credibility and reputation of enterprises which is the important factor for partner selection. Therefore, trust and reputation plays an important role in the operation of the virtual enterprises. With the Amerian financial crisis, the European debt crisis and international trade protectionism, the loss of credits between countries spread to the enterprises. The increase of uncertain factors led to the worse situation of losing credibility in global manufacturing industry. Trust crisis and credit deterioration in the cooperative enterprises have highlighted the risk of virtual enterprise.
TRUST EVALUATION MODEL USING QFD METHOD
QFD is applied in a wide variety of services, consumer products, military needs and emerging technology products. The technique is also used to identify and document competitive marketing strategies and tactics. We use QFD to trust evaluation of partner in virtual enterprise as shown in Fig. 1. In this hypothesis, where is:
| • | Si (i = 1,2,…, m) stands for the Trust improvement demands and wi is the weight of Si (Pan, 2001) |
| • | Q is the autocorrelation matrix of Si, qil ( i, lε[1, m] ) is the correlation coefficient of Si and Sl |
| • | P is autocorrelation matrix of improvement measures ej , jε[1, n], Pil is the correlation coefficient of Si and Sl as shown in Table 1, i, lε[1, n] |
| • | R is relationship matrix of Si and ej, rij is the correlation coefficient of Si and ej, i, jε[1, n] |
| • | D is the column vector of di (i = 1, 2,…, m) which is the current trust improvement |
| • | D* is the column vector of di* (iε[1, m]), target trust improvement value whose grades shown in Table 2 |
| • | C is the factors affecting on trust, cil (i, l = 1, 2…, n) is the correlation coefficient of ei and cl. |
| Hj is the trust improvement rate, Hj = (λj-μj)/μj, μj is the current value, λj is the improved trust value |
PROJECT SCHEDULE OPTIMAL MODEL BASE ON REPUTATION
We regard reputation as a positive incentive of partner selection and profits distribution of virtual enterprise, so that increase and decrease the credibility would have incentives and penalties to business interests. Based on the above ideas, project schedule model of virtual enterprise can be setup according to the principle of maximizing long-term interests which reflects the reputation as the center of modern management concepts.
| Fig. 1: | QFD House of trust evaluation |
| Table 1: | Trust evaluation set |
| Table 2: | Trust evaluation grades |
Suppose virtual enterprises should complete m tasks to produce n products within the planning period [1, T], di (t) is the order quantity of product i in day t. It is known the wij stands for unit production capacity of product i needs the resources j, αi is cost to keep a contract , βi is the cost to break a contract , i = 1, 2,…, n. The enterprise available capacity of day t is Cj (t), j = 1, 2,…, m, t = 1,2,…, T. The initial storage capacity of product i is Ii, Ii<0 indicates less production, i = 1, 2,…, n. Establishment of incentive mechanisms based on Ii will play a positive role for the maintenance of enterprise credit. Assuming a dynamic credit of enterprise Vi is with proportional to Ii, there are: Vi = KxIi+N (Shi-Hua and Yong-Lin, 2000).
Storage capacity can be expressed as Ii = (Vi-N)/K, K>0. K is the proportionality constant, N is fixed constant:
| (1) |
The planned production capacity of product i in day t is pi (t), i = 1, 2,…, n, t = 1, 2,…, T, then keep or break contract plan model (P0) as follows, namely, how full use of the effective resources to breach the minimum amount of punishment during planning period.
Since, the objective function of the above plan (P0) is nonlinear, can not use ordinary mathematical programming method for solving.
| Table 3: | Business factors affecting on keeping or breaking contract |
Let xi (t) and yi (t) were the more or less on the production of product i, taking into account the role of credit rating maintained, there:
| (2) |
Definition: xi (0), yi (0) are yield or less production of product i at end of the last plan period, apparently Ii = xi (0)-yi (0), according to the significance of each variable, the following recurrence formula:
| (3) |
Take x, y as variables, the P0 can be transformed into the following standard form linear program (P):
| (4) |
Solving the linear problem, the production of i in the day t, namely P (t), can be determined by the formula:
| (5) |
The key to solving the problem is as follows:
| Step 1: | Determination of αi and βi the business factors affecting on keeping or breaking contract of Virtual Enterprise are shown as Table 1. Let αij is the j kind of cost to keep contract of product i, Wij is the weight of stocking cost in advance which can be given by AHP method or empirical formula. The calculation of the cost to break contract is similar to keep it, there are: |
| (6) |
In general, αi and βi are proportional to the price of contract. To simplify, let Pi is contract price for product i, α and β are respectively for early and tardiness factor, αi and βi can be solved using the following formula:
| (7) |
| Step 2: | The constraint number of original problem (P0) is mxT, when m is large, the scale of the problem becomes too large to solve |
With the increasing of uncertainties in the world, whether keeping contract or not depends not only on the task of capacity constraints but also on exchange rates, raw material costs. Removing the force majeure, such as trade regulation, trade protection, only the real capacity constraints of those “bottleneck task” determine whether a company keep contract or not. The “bottleneck task” refers to the task with longer completion time on the critical path, it can be pointed by setting a time threshold. The task whose occupied time longer than the threshold is the “bottleneck task” (Pan, 2004).
Suppose Jit is the “bottleneck task” located at time t which constituted by tasks taking the j kind of resource to produce the i products, i = 1, 2,…, n, t = 1, 2,…, T. RP is a plan with relaxation and constrained only by jit.
Based on the above analysis, we give the following virtual enterprise project schedule algorithm:
| • | According to the nature of virtual enterprise, manufacturing process framework and its sub-classification process are setup |
| • | Taken the centralized control tasks in the framework to compose a directed graph G for task network |
| • | Identify the critical path in the directed graph |
| • | Identify the “bottleneck task” on the critical path, namely {jit} |
| • | Solving RP only constrained by{jit} and get the optimal solution x *, y * and p * |
| • | Check all constraints on the feasibility of p*, if all constraints can be satisfied, then p* is the optimal solution, stop; Otherwise, all the constraints not meet are added to the Relaxed Plan (RP) and will turn (Table 3) |
CONCLUSION
In this study, we presented our experience on optimizing project schedule of virtual enterprise. The trust evaluation applied QFD analysis to analyze the factors of trust crisis and credit deterioration in the countries, then cooperative enterprises and afterwards inside their own enterprise. We obtained preliminary results on trust crisis in virtual enterprise is of the increase of uncertain factors. Our results based on truest evaluation indicate that decision-making model should be based on reputation and trust incentive mechanisms should be built ASAP to avoid the risk of partner selection in virtual enterprise.
ACKNOWLEDGMENT
The authors wish to thank the Humanities and Social Sciences Foundation of Chinese Education Ministry (2324), Science and Technology Major Soft Science Foundation (2008GXS1D042), Ningbo Natural Science Foundation (2011A610106), Zhejiang philosophy and social sciences key research base-port modern service and creative industries research center, under which the present work was possible.