Advances in Production Engineering & Management Volume 12 | Number 2 | June 2017 | pp 173-184 https://doi.Org/10.14743/apem2017.2.249 ISSN 1854-6250 Journal home: apem-journal.org Original scientific paper Vehicle scheduling based on plant growth simulation algorithm and distribution staff behavior Ren, X.Y.a, Kong, Z.F.a, Liang, W.C.a*, Li, H.C.a, Zhou, X.Y.a aHebei University of Engineering, School of Management Engineering and Business, Handan, Hebei, P.R. China A B S T R A C T A R T I C L E I N F O Considering the distribution staffs satisfaction and fuzzy characteristics of customers' time window, the paper makes fuzzy description and measurement of staff satisfaction from such three aspects as salary and welfare, working strength and harmonious cooperation. Taking the minimum total logistics distribution cost and maximum distribution staff satisfaction as the objective functions, the study constructs the logistics distribution vehicle scheduling model considering the staff satisfaction and fuzzy time window. The plant growth algorithm is designed to solve the logistics distribution vehicle scheduling model. The simulation results show that the proposed the model succeeds in improving the customer satisfaction and the distribution efficiency, and that the optimization algorithm is feasible and effective. © 2017 PEI, University of Maribor. All rights reserved. Keywords: Vehicle scheduling Logistics distribution Staff satisfaction Plant growth simulation algorithm *Corresponding author: 12357414@qq.com (Liang, W.C.) Article history: Received 7 March 2017 Revised 15 May 2017 Accepted 16 May 2017 1. Introduction In logistics distribution, it is of critical importance for logistics distribution enterprises to provide timely and satisfactory service to customers. The service quality hinges on the psychology and behavior of distribution personnel who are in direct contact with customers. Such personnel normally work in high-intensity environments for long hours and often handle multiple tasks at the same time. This is particularly true for logistics distribution staff. Besides delivering goods to the designated places, they have to guarantee that the goods are intact and fresh. Whether they can provide timely and satisfactory distribution service to customers largely depends on their satisfaction with job. In light of this, designers of logistics distribution vehicle scheduling plans must take the job satisfaction of distribution staff into account. At present, scholars at home and abroad often put emphasis on quantitative factors like cost and time during the modelling and structural research of vehicle scheduling problem, aiming to minimize the distribution cost of the distribution center under certain constraints (volume, service time or distance), such as shortest route [1, 2, 3], minimum distribution cost [4, 5] and minimum number of vehicles [6]. Erdogan S. (2012) et al. proposed the concept of green vehicle scheduling [7]. Probing into a number of vehicle scheduling problems, Victor (2013) and Tas (2014) put forward several algorithms to solve the corresponding problems [8,9]. Li Juan and Yu Guoyin (2015) established the vehicle scheduling model with the goal of achieving the optimal comprehensive cost, which consists of fixed cost of vehicles, fuel cost and driver pay [10]. Sun Qiongling (2010) and Zhao Rui (2015) introduced customer satisfaction index into logistics distribution vehicle scheduling [11, 12]. Targeted at vehicle failure problem in logistics distribution, Cao Qingkui and Shao Songjuan (2016) constructed the emergency vehicle scheduling model in 173 Ren, Kong, Liang, Li, Zhou consideration of the distribution subject awareness [13]. Trancossi M (2016) et al. designed an innovative off-road hybrid vehicle [14]. The research on job satisfaction started early and yielded fruitful results in foreign countries. For instance, Locke (1970) classified the influencing factors of job satisfaction [15] nearly half a century ago. In recent years, foreign scholars have conducted in-depth studies on job satisfaction. Brian N. Rutherford (2014) et al. examined gender influence on job satisfaction and developed customized measures to promote the job satisfaction of the ever-increasing proportion of female employees [16]. Jarrod M. Haar (2014) et al. studied how employees in seven different cultural regions strike a balance between work and life, such as Malaysia, China and New Zealand, to name but a few [17]. Yam B. (2014) et al. suggested that information technology and communication skill are two driving factors of job satisfaction [18]. Scholars also explored the job satisfaction in different professions, like teachers and agents [19, 20]. As above, the research on job satisfaction has become more and more detailed. With the aid of new technologies and new theories, an increasing number of scholars have pursued how the interaction of various factors and different professions affect job satisfaction. With the rapid development of logistics industry in recent years, the standards on logistics service have become more stringent. In distribution activities, the psychology and behavior of distribution staff exert a great impact on the quality of distribution service, which in turn determines the enterprise reputation. Against this backdrop, it is better aligned with the actual situation to develop an optimization model for logistics distribution vehicle scheduling in the following steps: combine the research on the basic theory and behavioral pattern of logistics distribution, establish the optimization model for logistics distribution vehicle scheduling based on the job satisfaction of distribution staff, conduct optimization processing based on multiple objective functions in the model, merge the functions into a single objective function by linear weighting, and apply plant growth simulation algorithm to the solution of the problem. Following these steps, this paper presents a logistics distribution vehicle scheduling model based on distribution staff satisfaction in consideration of the actual behavior characteristics of distribution staff, and in light of the reasonable and valid distribution mode. The proposed model facilities logistics enterprises to improve their capability and performance of logistics distribution in an ever-increasingly competitive market. 2. Mathematical model for vehicle scheduling 2.1 Problem description From the perspective of distribution staff satisfaction, this paper summarizes the objectives of vehicle scheduling problem based on staff satisfaction, aiming to achieve the following traditional objectives: • Minimum vehicle transportation cost; • Maximum staff satisfaction; • Lowest satisfaction value of customer satisfaction. In this paper, the minimum requirement of customer satisfaction is set as 9. Fuzzy function of staff satisfaction We set ael,ae2, ---as the states of satisfaction of staff member e in the enterprise described in n indexes, and the satisfaction yeof staff member e in the enterprise is: where /(a)is the satisfaction function of staff and Aet is the weight coefficient [21]. If a > 0, then f{a) >0; if a<0, then /(a)<0;0+œ] (3) Le e [L Fig. 1(c) shows the fuzzy description of distribution staff harmonious cooperation satisfaction. The fuzzy membership function of logistics distribution staff satisfaction described by the harmonious cooperation Ee can be expressed as: 1, Eee[0,ELe] SE(ße) = {(Ee ~ELe)/(EHe ~ELe),Ee e(ELe,EHe) 0, Ee e [EHe,+œ] (4) Determination of element weight in staff satisfaction fuzzy description The staff satisfaction function Ye is fuzzily described by salary and welfare P, working strength L and harmonious cooperation E, which is: N n N N N ye = YJYu = ZÀp "Sp(Pe) + ZÀL "Sl(Le) + Z Ae "SE(Ee) (5) e=l 1=1 e=l e=l e=l Advances in Production Engineering & Management 12(2) 2017 175 Ren, Kong, Liang, Li, Zhou In staff satisfaction function Ye, A describes the different degrees of emphasis on each index among the distribution staff. In other words, A balances the three elements required for the fuzzy description of staff satisfaction. As mentioned above, the optimization scheduling model based on distribution staff satisfaction is required to achieve the maximum distribution staff satisfaction. N n N N N maxYe = max ^ ^ Aelf(ael) = max ^ Ap • SP(Pe) + ^ AL • S¿(Le) + ^ AE • SE(Ee) e = l 1 = 1 Le = l e = l e = l (6) Fuzzy description of customer satisfaction First, distribution punctuality is adopted to fuzzily describe customer satisfaction. The time period [Tzi,Twi] represents the service time period accepted by customers; the time period [Tzi,Twi\ represents the service time period that satisfies customers the most. Second, Ut(T{) is used to express the fuzzy satisfaction of customers and the fuzzy membership function of logistics distribution satisfaction of customer i as described by the distribution punctuality 7V The parameter can be expressed as: 0 (Tt -Tzi)/(TZi -Tzt) 1 (TWi -Ti)/(Twi -TW{) 0 Urft) = O^Ti^TZi TziTWi In this research, logistics distribution must firstly fulfil the goal of the minimum requirement of customer satisfaction. In light of the fuzzy satisfaction of customers, the goal is expressed as min UT(Ti) = 0. 2.2 Parameter variable and correlation analysis Parameter variable We use i = 0 to represent the distribution center and i = 1,2, -^m to represent the m customers. The relevant parameters and variables are defined as follows. m: the m customers of distribution, denoted as 1,2,3,•••, i, •••,_/, ••• ,m ; N : the N staff members, denoted as 1,2,3, —, e, •••, N; n: the n staff satisfaction indexes, denoted as 1,2,3, •, I, •••, n; K: the K distribution vehicles, denoted as 1,2, •••, k; qt: the distribution quantity demanded by customer i; vt: the measurement of distribution goods of customer i; Qk: the maximum loading capacity of the kth distribution vehicle; Vk: the maximum loading volume of the kthdistribution vehicle; FCk: the fixed cost of the kth distribution vehicle, i.e. the marginal cost incurred by the addition of a new vehicle; DCk: the unit distance cost of the kthdistribution vehicle;^: the time when the distribution vehicle reaches customer i; t^: the transportation time, i.e. period of time for the distribution vehicle to move from customer i to customer j; t[: the service (unloading) time of the distribution vehicle at customer i, hereinafter referred to as the service (unloading) time; v: the average running speed of the distribution vehicle; d^: the distance between customer i and er j; L: the driving distance reaching the compensation point; Lmax: the maximum running distance of the distribution vehicle in a single distribution activity; S: the compensation cost coefficient of distribution vehicle beyond the additional compensation point;^: the basic salary of distribution staff per unit time during normal working hours; k2: the pay per unit time during overtime; t0: the normal working hours of staff members; t: the actual working time of distribution staff members, including transportation time t^ and service (unloading) time t-; ts: the time for the distribution vehicle to depart from the distribution center; tb: the time for the distribution vehicle to return to the distribution center. 176 Advances in Production Engineering & Management 12(2) 2017 Vehicle scheduling based on plant growth simulation algorithm and distribution staff behavior Establishment of distribution optimization model Objective functions: m m K K minC = YjDCk •diJ •XiJk + FCk + p& + ST(dij) (8) í=ij=ifc=i fc=i N n N N N maxye = max^^ Xelf(ael) = max[^ Ap • SP(Pe) + ^ XL • SL(Le) + ^ ÀE • f] e = li=l e = l e=l e=l minyT(r¿) = 9 Constraint conditions: •qij = 1,2,3, — , m; k = 1,2,3,— , rik = 0 or 1, i = 1,2,3, — ,m; k = 1,2,3, —,K Eq. 8 is the minimum total logistics distribution cost, Eq. 9 is the maximum distribution staff satisfaction, and Eq. 10 is the minimum requirement of customer satisfaction of logistics distribution service. Constraint conditions in Eqs. 11 and 12 describe the capacity limitation of distribution vehicle. Specifically, Eq. 11 means the goods should not exceed the maximum loading weight of the vehicle and Eq. 12 means each vehicle is able to hold the goods in the distribution activities. Constraint conditions in Eqs. 13 and 14 depict the running limitations of the distribution vehicle. Specifically, Eq. 13 means each customer point is served only once by one distribution staff member and one distribution vehicle, and Eq. 14 means each staff member and vehicle leaves from the distribution center, and returns to the distribution center after completing the Advances in Production Engineering & Management 12(2) 2017 177 Ren, Kong, Liang, Li, Zhou distribution. Constraint conditions in Eqs. 15 to 18 present the time limitations of distribution vehicle. Specifically, Eq. 15 describes the transportation time, i.e. the time for the vehicle to move from one customer to the next customer; Eq. 16 illustrates the arrival time at the next customer, which equals to the sum of the arrival time and the service (unloading) time at the current customer, and the transportation time; Eqs. 17 and 18 represent the time for the vehicle to depart from or return to the distribution center, respectively; both of the two parameters are set as 0. Constraint conditions in Eqs. 19 and 20 show the working limitations of the distribution staff. Specifically, Eq. 19 specifies that the working time of the distribution staff should not exceed the maximum working time during each distribution; Eq. 20 stipulates that the running distance of the distribution vehicle should not exceed the maximum running distance during each distribution. Last but not the least, constraint conditions in Eqs. 21 and 22 determine the value of decision variables. Analysis of objective Junctions In distribution activities, the total cost is composed of distribution cost, salary of distribution staff and compensation cost. The distribution cost is further divided into the fixed cost of logistics distribution vehicle and running cost, which are expressed as: =FCkhk=lrik, •dij •Xijfr. It is assumed that the goods in the distribution are common goods of the same quality and the vehicles are moving at a constant speed. Consisting of basic salary and overtime salary, the salary of distribution staff is expressed in piecewise function of working time: C3 -p(t) klt' 1 - ío (23) (Mo + to), t>to 1 J Compensation cost: Whereas the working strength of distribution staff increases when the driving distance in a single distribution is excessively long, it is necessary to give distribution staff some compensation. In this research, the working strength of distribution staff is quantified based on the running distance of the distribution vehicle in one distribution activity. It is expressed as below. m m T(di]) = max[2j2j(di] • xi]k -i), 0] (24) ¿=17=1 The compensation cost function can be expressed as follow: K K m m Q = T(dij) = S ^ •xiJk -i), 0] (25) k=1 k=1 ¿=17=1 Since the distribution staff satisfaction should be determined through quantification of salary and welfare, working strength and harmonious cooperation, the distribution staff satisfaction function can be written as follows if the harmonious cooperation is regarded as a constant: W n maxYe = max^^2eí/(aeí) = max e = l 1 = 1 r n N N ^ Ap • SP(Pe) + ^ V SlUJ + ^ h • SE(Ee) e = l e = l e = l WWW • Sp(Pe) + ^ Xh • S¿(Le) + ^ Ae • g (26) = max e=l e=l e=l 178 Advances in Production Engineering & Management 12(2) 2017 Vehicle scheduling based on plant growth simulation algorithm and distribution staff behavior 3. Configuration of vehicle scheduling by plant growth simulation algorithm 3.1 Optimized scheduling model based on plant growth simulation algorithm Inspired by the growth of plants in nature, the plant growth simulation algorithm (PGSA) is a type of optimization algorithm used to simulate the plant growth characteristics. Taking a certain plant as the object, the length of the main stem is denoted as M, the length of the branch is denoted as m, the number of growing points on the main stem is denoted as K, which can be expressed as Sm = (Smi, Sm2, ••• , SW), the auxin concentration of growing points is denoted as Pm = (Pmi, SPm2, ••• , fW), the number of growing points on the branch is denoted as q, which can be expressed as Sm = (Smi, Sm2, ••• , Smq), the auxin concentration at growing points is denoted as Pm = (Pmi, SP¡ m2, Pmq). The auxin concentration at each growing point on the main stem and branch of this plant can be calculated according to the following formulas: Pmí — PMJ — f(.x0)-f(Sml) £f=l(/(*o) - f(Smù)+ I-=1(/(*o) - f(Smi)) f(.x0)-f(Sml) rf=i (/(* o) - f(Sm])) + SJ=1(/(X0) - f(Sm])) mí + — 1 (27) (28) (29) 1=17=1 3.2 Design and implementation of algorithm flow The X is used to represent the bounded closure on the Rn and the f(x) is used to represent the objective function. Thus, the flow and implementation of the PGSA are shown in Fig. 2. Determine the step of PGSA, set the initial seed node as, initialize As,m= a», /;„,„= /(a»), a= as _*_^_ Generate random number rando within the range from 0 to 1. If ^pm. SVr_ The optimal solution is xm\n and the optimal value is Fmin. Fig. 2 The flow and implementation of the PGSA Advances in Production Engineering & Management 12(2) 2017 179 Ren, Kong, Liang, Li, Zhou 4. Simulation analysis based on the PGSA 4.1 Data collection In light of the lack of standard testing dataset for the research on vehicle scheduling problem, this paper designs a concrete algorithm instance. The objective of initial scheduling plan is to minimize logistics distribution cost and maximize customer satisfaction. The demand information parameters and time window constrains for 22 customer points are selected randomly by a computer. Table 1 and Table 2 list the coordinates of the customer points and the demand information. Table 1 Customer point coordinates and demand information Number Customer location Quantity Service (unloading) Time window X coordinate Y coordinate /piece ti /min 0 26 33 — — — 1 32 19 21 25 [7:00-8:00] 2 18 35 23 30 [9:00-11:45] 3 34 54 30 40 [7:00-9:00] 4 42 51 19 20 [8:00-10:45] 5 24 28 13 16 [7:00-11:45] 6 17 26 17 20 [7:00-8:45] 7 37 11 21 26 [13:00-18:00] 8 54 20 13 18 [14:00-16:45] 9 30 27 11 14 [12:00-18:00] 10 48 35 29 38 [7:00-8:45] 11 51 48 19 38 [14:00-16:45] 12 25 19 9 10 [9:00-11:45] 13 26 26 25 35 [7:00-8:45] 14 11 24 15 20 [7:00-8:45] 15 36 23 11 16 [15:00-17:45] 16 30 36 13 15 [13:00-17:00] 17 54 18 23 30 [17:00-18:00] 18 48 40 21 24 [14:00-17:45] 19 18 54 12 18 [8:00-18:00] 20 42 29 15 22 [9:00-18:00] 21 24 41 21 25 [7:00-15:45] 22 53 42 24 30 [13:00-18:00] Table 2 Distance between each customer point and the distribution center 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 0 0 18 30 15 29 9 16 28 35 8 27 32 20 8 18 15 8 35 25 28 19 10 30 1 18 0 25 40 40 18 19 12 30 14 29 40 9 13 25 8 20 29 28 40 16 28 37 2 30 25 0 32 30 13 11 32 45 20 36 40 23 17 15 23 25 40 34 23 29 11 40 3 15 40 32 0 9 28 35 48 43 30 26 22 39 32 40 39 24 45 28 20 30 18 27 4 29 40 30 9 0 34 40 45 39 30 23 12 38 34 48 30 22 39 20 26 28 29 18 5 9 18 13 28 34 0 10 26 40 12 30 40 19 8 18 23 20 43 30 38 35 32 43 6 16 19 11 35 40 10 0 35 56 23 45 57 34 15 19 34 28 48 42 47 46 42 45 7 28 12 32 48 45 26 35 0 23 25 35 48 25 35 48 22 36 38 47 69 34 68 48 8 35 30 45 43 39 40 56 23 0 34 35 55 35 43 64 25 38 7 34 57 34 40 44 9 8 14 20 30 30 12 23 25 34 0 23 34 12 7 23 10 19 34 45 39 23 35 35 10 27 29 36 26 23 30 45 35 35 23 0 24 34 35 46 38 33 37 14 58 24 57 27 11 32 40 40 22 12 40 57 48 55 34 24 0 47 57 89 33 43 40 12 35 24 30 12 12 20 9 23 39 38 19 34 25 35 12 34 47 0 12 16 22 34 41 56 43 20 22 38 13 8 13 17 32 34 8 15 35 43 7 35 57 12 0 23 34 23 44 34 54 23 41 39 14 18 25 15 40 48 18 19 48 64 23 46 89 16 23 0 29 34 54 56 78 66 43 76 15 15 8 23 39 30 23 34 22 25 10 38 33 22 34 29 0 34 34 35 56 24 35 40 16 8 20 25 24 22 20 28 36 38 19 33 43 34 23 34 34 0 40 35 34 32 16 38 17 35 29 40 45 39 43 48 38 7 34 37 40 41 44 54 34 40 0 34 98 34 56 47 18 25 28 34 28 20 30 42 47 34 45 14 12 56 34 56 35 35 34 0 45 32 38 15 19 28 40 23 20 26 38 47 69 57 39 58 35 43 54 78 56 34 98 45 0 47 32 56 20 19 16 29 30 28 35 46 34 34 23 24 24 20 23 66 24 32 34 32 47 0 32 34 21 10 28 11 18 29 32 42 68 40 35 57 30 22 41 43 35 16 56 38 32 32 0 45 22 30 37 40 27 18 43 45 48 44 35 27 12 38 39 76 40 38 47 15 56 34 45 0 180 Advances in Production Engineering & Management 12(2) 2017 Vehicle scheduling based on plant growth simulation algorithm and distribution staff behavior 4.2 Parameter setting of the model and the algorithm Model parameters In this research, the number of distribution vehicles is set as 4, and the nominal load, average driving speed d, per kilometer driving cost, and the maximize driving distance in a single distribution of each vehicle are set as 2,000 kg, 30km/h, 2.5 yuan, and 200km, respectively. Any distribution activity beyond the maximum distance will incur compensation cost. The salary of distribution staff in working hours is 15 yuan/h and the overtime salary is 30 yuan/h. The details of the parameter setting are shown in Table 3. Table 3 Parameter setting of the model Parameter Meaning Value FCk The fixed cost of the distribution vehicle 1S0 DCK The unit distance cost of the distribution vehicle 2.S V The average driving speed of the distribution vehicle 30 L The driving distance reaching the compensation point 120 L-max The maximum distance in a single distribution activity 200 S The compensation cost coefficient of the distribution vehicle 100 k1 The basic salary per unit time of distribution staff during normal working hours is k2 The pay per unit time during the overtime 30 to The normal working hours of staff members 6 t-max The maximum working hours of distribution staff 10 9 The minimum requirement of customer satisfaction 0.75 K The harmonious cooperation of distribution staff 0.8 Algorithm parameters This paper adopts the PGSA to solve this model. The main parameter settings of the algorithm are: population size is 30, crossover probability is 0.8, mutation probability is 0.05, number of generation is 100, and step is 1. 4.3 Optimization of the model One deputy general manager, two scheduling managers, five distribution staff members and two experts are invited to weigh the importance of salary Pe, working strength Le and harmonious coopetition Ee. Each person determines the weight of Ap, Al and Ae, and the final results are: the weight of salary is 0.51, the weight of working strength is 0.27 and the weight of harmonious coopetition is 0.22. Without considering staff satisfaction, the logistics distribution activity aims to minimize the total distribution cost and meet the minimum requirement of customer satisfaction in the same distribution activity. Irrespective of customer satisfaction, we can obtain the following distribution plan (Table 4) by solving the original scheduling problem. In this plan, the total distribution cost is 2,878 yuan and the total distribution distance is 443km. The average satisfaction of all customers is 77.32 % and the average satisfaction of distribution staff is 58.63 %. The distribution route and vehicle scheduling results are shown in Fig. 3 and Table 5. Table 4 Vehicle scheduling plan without considering distribution staff satisfaction Vehicle k Optimal route 1 0-2-19-3-21-0 2 0-13-5-6-14-12-7-1-9-0 3 0-15-17-8-10-20-0 4 0-16-4-11-22-18-0 Advances in Production Engineering & Management 12(2) 2017 181 Ren, Kong, Liang, Li, Zhou Fig. 3 Distribution vehicle scheduling route without considering distribution staff satisfaction Table 5 Vehicle scheduling results without considering distribution staff satisfaction Distribution route Fixed cost /yuan Transportation cost / yuan Labor cost during normal working time/yuan Labor cost during overtime/ yuan Total cost / yuan Average customer satisfaction, (%) Distribution staff satisfaction, (%) Route1 150 252.5 78.75 0 481.25 77.16 58.2 Route2 150 302.5 90 24 666.5 80.63 53.41 Route3 150 317.5 90 9 1266.5 76.4 48.73 Route4 150 235 78.75 0 463.5 75.1 74.18 Considering distribution staff satisfaction, this paper adopts Matlab7.1 coding to realize the algorithm based on the proposed vehicle scheduling optimization model and plant growth simulation algorithm. The solution is carried out on a laptop of Intel(R) Core(TM) i3-380M CPU 2.53 GHz and 4 GB memory. The average time consumption is 13.5 s and 4 distribution routes are obtained. The optimal scheduling plan of distribution vehicle in this instance is shown in Table 6. Table 6 Vehicle scheduling plan considering distribution staff satisfaction Vehicle k_Optimal route 1 0-6-14-2-19-21-16-0 2 0-3-4-11-22-18-0 3 0-10-8-17-20-0 4 0-5-13-12-1-7-15-9-0 For this plan, the optimized route and the scheduling results are shown in Fig. 4 and Table 7, respectively. We can see that the total distribution distance of this logistics activity is 429km and the distribution cost is 3,126.05 yuan. The average satisfaction of all customers is 83.5 % and the average satisfaction of distribution staff is 91.93 %. 182 Fig. 4 Logistics distribution vehicle scheduling route considering distribution staff satisfaction Advances in Production Engineering & Management 12(2) 2017 Vehicle scheduling based on plant growth simulation algorithm and distribution staff behavior Table 7 Vehicle scheduling results considering distribution staff satisfaction Distribution Fixed cost Distribution cost (yuan) Labor cost during normal working time/yuan Labor cost during over-time/ yuan Total cost /yuan Average customer satisfaction (%) Distribution staff satisfaction (%) Routel Route2 Route3 Route4 150 150 150 150 322.5 220 305 225 90 82.05 88.05 80.55 12.9 0 0 0 1475.4 452.05 743.05 455.55 82 89 85 78 92.6 91.1 93.8 90.2 4.4 Contrastive analysis of results According to the contrastive analysis of the optimization results and the vehicle scheduling results without considering distribution staff satisfaction, we have the following findings: • Compared with the scheduling plan without considering distribution staff satisfaction, the customer satisfaction in the proposed distribution plan has improved significantly, which is less than 2 0 %; • The optimized distribution plan has improved the distribution efficiency significantly as it costs less time to complete the same logistics distribution activity and lower maximum time consumption of a single scheduling route than the original distribution plan; • The logistics cost of the optimized scheduling plan is 8.6 % higher than that of the original plan. It can be inferred from the previous analysis that the distributions staff satisfaction increases at the expense of the distribution cost. With the improvement of distribution staff satisfaction, however, the time consumption of the same distribution activity in the proposed plan is significantly less than that in the original plan. The tremendous improvement of the distribution efficiency and customer satisfaction would contribute to customer retaining and reduction of customer loss. Moreover, the improvement of distribution staff satisfaction also boosts the working enthusiasm and the sense of ownership of the staff members and reduces staff turnover rate. The results are conducive to the long-term development of enterprises. 5. Conclusion For the purpose of establishing the logistics distribution vehicle scheduling model based on distribution staff satisfaction and time window, this paper makes fuzzy description of the variables of distribution staff satisfaction, e.g. salary and welfare, working strength and harmonious cooperation, and takes them as the constraints. Then, the author sets up the plant growth simulation algorithm model of the distribution vehicle scheduling, and obtains the scheduling optimization plan and scheduling results through simulation analysis. The new scheduling plan manages to improve the satisfaction degree and distribution efficiency at the expense of higher distribution cost. The subsequent research will deal with the uncertainties in the actual situation, such as bad weather, road blockage, vehicle failure, changes of customers' time window, etc. Acknowledgment The research of this paper is made possible by the generous support from Natural Science Foundation of Hebei Province (F2014402040), Social Science Foundation of Hebei Province (HB16GL026), Research Projects of Innovative Talents Training Fund of Hebei Province (A2016001120), Scientific Research Planning Projects of Education Department in Hebei Province (SD161009) and Key Research Project of Humanity and Social Science Research of Education Department in Hebei Province (ZD201442). Advances in Production Engineering & Management 12(2) 2017 183 Ren, Kong, Liang, Li, Zhou References [1] Jiang, D., Yang, X., Du, W., Zhou, X. (1999). A study on the genetic algorithm for vehicle routing problem, Systems Engineering - Theory & Practice, Vol. 19, No. 6, 40-45, doi: 10.3321/j.issn:1000-6788.1999.06.007. [2] Lang, M.X., Hu, S.J. (2002). Study on the optimization of physical distribution routing problem by using hybrid genetic algorithm, Chinese Journal of Management Science, Vol. 10, No. 5, 51-56, doi: 10.3321/j.issn:1003-207X. 2002.05.011. [3] Berger, J., Barkaoui, M. (2004). A parallel hybrid genetic algorithm for the vehicle routing problem with time windows, Computers & Operations Research, Vol. 31, No. 12, 2037-2053, doi: 10.1016/S0305-0548(03)00163-1. [4] Gong, Y.C., Guo, X.F., You, X.L., Zhao, Y. (2004). Solving the vehicle routing and scheduling problems by genetic algorithms, Mathematics in Practice and Theory, Vol. 34, No. 6, 93-97, doi: 10.3969/j.issn.1000-0984.2004. 06.017. [5] Hwang, H.-S. (2002). An improved model for vehicle routing problem with time constraint based on genetic algorithm, Computers & Industrial Engineering, Vol. 42, No. 2-4, 361-369, doi: 10.1016/S0360-8352(02)00033-5. [6] Renaud, J., Boctor, F.F. (2002). A sweep-based algorithm for the fleet size and mix vehicle routing problem, European Journal of Operational Research, Vol. 140, No. 3, 618-628, doi: 10.1016/S0377-2217(01)00237-5. [7] Erdogan, S., Miller-Hooks, E. (2012). A green vehicle routing problem, Transportation Research Part E: Logistics & Transportation Review, Vol. 48, No. 1, 100-114, doi: 10.1016/j.tre.2011.08.001. [8] Pillac, V., Gendreau, M., Gueret, C., Medaglia, A.L. (2013). A review of dynamic vehicle routing problems, European Journal of Operational Research, Vol. 225, No. 1, 1-11, doi: 10.1016/i.eior.2012.08.015. [9] Ta§, D., Jabali, O., van Woensel, T. (2014). A vehicle routing problem with flexible time windows, Computers & Operations Research, Vol. 52, Part A, 39-54, doi: 10.1016/i.cor.2014.07.005. [10] Li, J., Yu, G. (2015). Study on chaotic firefly algorithm for vehicle scheduling problem with minimal comprehensive cost, Logistics Technology, No. 4, 180-183, doi: 10.3969/i.issn.1005-152X.2015.04.055. [11] Sun, Q.-L., Wang, L. (2010). A Study on domestic business enterprises' vehicle scheduling problems based on satisfaction of customers, Logistics Sci-tech, Vol. 33, No. 6, 18-21, doi: 10.3969/i.issn.1002-3100.2010.06.007. [12] Zhao, R., Hu, X., He, H. (2015). E-commerce logistics distribution routing optimization considering customer satisfaction, Journal of Shanghai Maritime University, No. 3, 64-70, doi: 10.13340/i.ismu.2015.03.011. [13] Cao, Q.-K., Shao, S.-J., Ren, X.-Y. (2016). Study on the rescue vehicle scheduling considering the perception of the distribution subject, Logistics Sci-tech, Vol. 39, No. 2, 1-3, doi: 10.3969/i.issn.1002-3100.2016.02.001. [14] Trancossi, M., Pascoa, J.C., Xisto, C.M. (2016). Design of an innovative off road hybrid vehicle by energy efficiency criteria, International Journal of Heat and Technology, Vol. 34, No, 2, 387-395, doi: 10.18280/iiht.34Sp0228. [15] Locke, E.A. (1970). Job satisfaction and iob performance: A theoretical analysis, Organizational Behavior & Human Performance, Vol. 5, No. 5, 484-500, doi: 10.1016/0030-5073(70)90036-X. [16] Rutherford, B.N., Marshall, G.W., Park, J.-K. (2014). The moderating effects of gender and inside versus outside sales role in multifaceted iob satisfaction, Journal of Business Research, Vol. 67, No. 9, 1850-1856, doi: 10.1016/i.ibusres.2013.12.004. [17] Haar, J.M., Russo, M., Sune, A., Ollier-Malaterre, A. (2014). Outcomes of work-life balance on iob satisfaction, life satisfaction and mental health: A study across seven cultures, Journal of Vocational Behavior, Vol. 85, No. 3, 361373, doi: 10.1016/i.ivb.2014.08.010. [18] Limbu, Y.B., Jayachandran, C., Babin, B.J. (2014). Does information and communication technology improve iob satisfaction? The moderating role of sales technology orientation, Industrial Marketing Management, Vol. 43, No. 7, 1236-1245, doi: 10.1016/i.indmarman.2014.06.013. [19] Si, H.Y., Wang, Y., Yin, F.F. (2014). Effects of call center representative's psychological capital on work stress and iob satisfaction, Chinese Journal of Ergonomics, Vol. 20, No. 1, 41-44, doi: 10.13837/i.issn.1006-8309.2014.01.009. [20] Gagliano, A., Nocera, F., Detommaso, M., Evola, G. (2016). Thermal behavior of an extensive green roof: Numerical simulations and experimental investigations, International Journal of Heat and Technology, Vol. 34, No. 2, 226-234, doi: 10.18280/iiht.34S206. [21] Chen, P. (2001). Staff satisfaction model analysis and application, Industrial Technology & Economy, No. 3, 80-81, doi: 10.3969/i.issn.1004-910X.2001.03.036. [22] Yan, Z.-G. (2007). Management on the iob satisfaction of the third-party logistics company, Journal of Guangdong Communications Polytechnic, Vol. 6, No. 2, 70-72, doi: 10.3969/i.issn.1671-8496.2007.02.024. [23] Xu, L. (2012). Study on logistics enterprise human resource management based on employee satisfaction, Logistics Technology, Vol. 31, No. 8, 177-179, doi: 10.3969/i.issn.1005-152X.2012.08.061. [24] Wang, Y. (2013) Empirical study on employee satisfaction of logistics consignor enterprises: The Example of Xi'an, Logistics Technology, Vol. 32, No. 5, 348-349, doi: 10.3969/i.issn.1005-152X.2013.05.110. [25] Sun, L.-H., Huang, F., Yuan, X.-F., Shui, H.-H. (2015). The evaluation research of iob satisfaction among logistics employee in we-commerce company, Chinese Journal of Ergonomics, Vol. 21, No. 4, 30-35, doi: 10.13837/i.issn. 1006-8309.2015.04.0007. [26] Tan, T.T., Tan, J.T. (2016). Research on the relationship between iob stressors and iob burnout of employees in logistics enterprises, Modern Business, No. 13, 71-74, doi: 10.3969/i.issn.1673-5889.2016.13.035. [27] Li, T., Wang, C.-F., Wang, W.-B., Su, W.-L. (2005). A Global optimization bionics algorithm for solving integer programming - Plant growth simulation algorithm, Systems Engineering - Theory & Practice, Vol. 25, No. 1, 76-85, doi: 10.3321/i.issn:1000-6788.2005.01.012. 184 Advances in Production Engineering & Management 12(2) 2017