1. A vehicle optimization scheduling method in a kitchen waste recovery process is characterized by comprising the following steps: the method comprises the following steps:

step1, constructing a fuzzy opportunity constraint planning model of the vehicle optimization scheduling problem in the kitchen waste recovery process based on a fuzzy credibility theory, and minimizing the total economic cost summarized by the departure cost, the travel path cost and the time cost of the vehicle as an optimization target;

step2, solving the vehicle optimized dispatching problem determined in Step1 in the kitchen waste recovery process by adopting an optimized dispatching method of a discrete artificial bee colony algorithm to obtain a recovery scheme before the kitchen waste recovery of the vehicle;

and Step3, recycling the kitchen waste for the customers according to the recycling scheme in Step2, and adjusting the recycling scheme in real time by adopting a rescheduling method so as to reduce the extra cost caused by service failure of part of the customers in the recycling scheme due to uncertainty of the kitchen waste.

2. The vehicle optimization scheduling method in the kitchen waste recovery process according to claim 1, characterized in that: the fuzzy opportunity constraint planning model constructed based on the fuzzy credibility theory in Step1 is as follows:

s.t.

where V ═ 0 ═ vu { (0 })₀，V₀1,2,3 …, n is the set of customer points, 0 represents the yard, n is the total number of customers; k {1,2, … m } is the set of available vehicles in the yard, m is the total number of vehicles; r is^kThe total number of vehicles required for the dispatch protocol; c is the fixed cost of vehicle departure; c. C₁Is the unit travel cost of the vehicle; c. C₂A waiting cost for early arrival of the vehicle; d_ijDistance from client i to client j;is the waiting time of vehicle k at customer i; q is the maximum load capacity of the vehicle;representing the amount of fuzzy kitchen waste of customer j, representing triangular fuzzy numbersA measure of credibility of Cr ∈ [0,1 ]]The larger Cr is, the greater credibility that the remaining load capacity of the vehicle can meet the fuzzy demand of the client is; α is a given preference value; b_0kIs the departure time of vehicle k; a is_0kTime for vehicle k to return to yard; e.g. of the type₀An open time window for the yard; l₀The closing time of the yard; t is t_ijA travel time for the vehicle from customer i to customer j;time of arrival of vehicle k at customer i;time of departure for vehicle k from customer i; s_iA service time for the vehicle at customer i; e.g. of the type_iThe start time of the time window is client i; l_iThe closing time of the time window is client i; x is the number of_ijkAs a decision variable, if vehicle k is 1 from customer i to customer j, otherwise it is 0; y is_ikFor decision variables, 1 if customer i is served by vehicle k, and 0 otherwise.

3. The vehicle optimization scheduling method in the kitchen waste recovery process according to claim 1, characterized in that: the optimized scheduling method based on the discrete artificial bee colony algorithm in Step2 comprises the following specific steps:

step2.1, encoding and decoding mode: an integer coding and decoding strategy based on customer arrangement is adopted, the number of the train yard is 0, the number of the customer is an integer larger than 0, and during coding, the customer arrangement without the train yard is adopted for coding;the individual i of the gen generation is decoded based on the arrangement of the clients by adopting a mode of routing first and then grouping according to the capacity of the vehicle and the time window constraint of the clients,is a pair ofA complete delivery scheme is obtained after decoding, S isThe length of S is more than or equal to 2+ n and less than or equal to 2n + 1;

step2.2, population initialization: generating SN initial populations with diversity and dispersity by adopting a random rule;

step2.3, bee hiring phase: placing the employment bees on the honey sources, allocating one employment bee for each honey source and evaluating the quality of the honey sources, wherein the number of the employment bees is SN, and the employment bees search for new honey sources nearby the allocated honey sources through a formula (15);

in the formula X_i，X_kRespectively, i-th and k-th honey sources, i ≠ k,represents a pair X_iAnd X_kPerforming on the basis of r [ a, b]Sequential interleaving of (1), r [ a, b ]]Representing a randomly generated length for controlling the distance of the crossover, wherein 1 ≦ a ≦ b ≦ n, evaluating the new honey source after the employed bee finds the new honey source, if X is the new honey source_kThe fitness value of the honey source is superior to that of the old honey source X_iThe fitness of (2) is X_kSubstitution of X_iOtherwise, the old honey source is reserved;

step2.4, bee stage observation: when all the employed bees complete the search of new honey sources, the observation bees calculate the probability p of each honey source being selected according to the information of the honey sources searched by the employed bees by the formulas (16) and (17), and then the honey sources are randomly selected by adopting the roulette method, wherein p in the formula (16)_iRepresenting the probability that the honey source i is selected, fitness_iThe fitness value of the ith honey source is expressed, and the fitness value is calculated as shown in formula (17), wherein f is shown in formula (17)_iExpressing an objective function value of the ith honey source, then searching the selected honey source according to the same search strategy as the employed bees, wherein the number of the observed bees is the same as the number of the employed bees, and the observed bees are SN;

step2.5, bee detection stage: after all observation bees finish the search of new honey sources, the selection judgment is carried out on all honey sources, if a certain honey source still cannot be updated after limit cycles, the food source X is_iWill be discarded, the hiring bee corresponding to this food source is transformed into a scout bee, which is replaced by a new food source generated by formula (18);

X′_i＝Swap(X_i,u,v) (18)

equation (18) is for X_iPerforming a swap neighborhood operation, randomly in sequence X_iTwo positions u and v are selected, and then the clients on the u position and the v position are interchanged;

step2.6, stop conditions: and setting the termination condition as the maximum iteration number, outputting an optimal path if the maximum iteration number is met, and otherwise, repeating Step2.3, Step2.4 and Step2.5 until the termination condition is met.

4. The vehicle optimization scheduling method in the kitchen waste recovery process according to claim 1, characterized in that: the rescheduling method in Step3 comprises the following specific steps:

step3.1, acquiring a recovery path of the current vehicle;

step3.2, judging whether the vehicle has the remaining space, if so, turning to Step3.3, otherwise, returning to the parking lot, and rescheduling the remaining customers which are not served by adopting a discrete artificial bee colony algorithm;

step3.3, whether the service of the customer distributed by the current vehicle is finished or not, if not, the service is continuously carried out according to the distributed path, and if the service is finished, the algorithm is ended.

5. The vehicle optimal scheduling method in the kitchen waste recovery process according to claim 3, characterized in that: the population size SN in step2.2 is set to 30, and the number of non-updates limit is 5.

Background

With the rapid development of economy, the acceleration of urban process and the rapid expansion of catering industry in China, the generation of kitchen waste is increased year by year. The kitchen waste has the dual attributes of pollutants and resources, can be changed into resources by changing waste into valuable when being properly treated, and pollutes the environment and is harmful to health when being improperly treated. According to statistics, kitchen waste of not less than 6000 million tons is generated in cities in China every year, and accounts for more than 50% of the total amount of the waste, wherein the daily output of kitchen waste of catering service units in cities such as Beijing, Shanghai and Shenzhen has broken through one thousand tons, and the daily output of kitchen waste of catering service units in other large and medium-sized cities is about several hundred tons, so that the huge yield of the kitchen waste can cause huge influence on catering industry and urban environment if the kitchen waste cannot be timely and effectively treated.

In the kitchen waste recovery process, the amount of the kitchen waste is often difficult to obtain accurately, and the kitchen waste generated in each large meal and beverage industry is related to a plurality of factors, such as holidays, weather and other factors. In addition, in actual kitchen waste recycling enterprises, reasonable scheduling is mostly lacked, unified management is not available, and even large-scale enterprises can only perform simple scheduling according to experience of workers.

In the whole kitchen waste recovery process, a kitchen waste recovery company has a plurality of recovery vehicles, the kitchen waste recovery company reaches a client within the time specified by the client to provide the service of recovering the kitchen waste for the client, and the problem of the optimized scheduling of the kitchen waste recovery process belongs to the NP-hard problem, so that the solving difficulty of the problem increases exponentially along with the increase of the number of the clients. For solving the problems, the traditional heuristic method cannot ensure the quality of the solution, and the mathematical programming method has better solving quality but longer time.

Disclosure of Invention

The invention provides a vehicle optimal scheduling method for a kitchen waste recovery process, which is used for obtaining an excellent solution of the vehicle optimal scheduling problem in the kitchen waste recovery process in a short time.

The technical scheme of the invention is as follows: a vehicle optimization scheduling method in a kitchen waste recovery process is characterized by comprising the following steps: the method comprises the following steps:

step1, constructing a fuzzy opportunity constraint planning model of the vehicle optimization scheduling problem in the kitchen waste recovery process based on a fuzzy credibility theory, and minimizing the total economic cost summarized by the departure cost, the travel path cost and the time cost of the vehicle as an optimization target.

And Step2, solving the vehicle optimized dispatching problem determined in Step1 in the kitchen waste recovery process by adopting an optimized dispatching method of a discrete artificial bee colony algorithm, and obtaining a recovery scheme before the kitchen waste recovery of the vehicle.

And Step3, recycling the kitchen waste for the customers according to the recycling scheme in Step2, and adjusting the recycling scheme in real time by adopting a rescheduling method so as to reduce the extra cost caused by service failure of part of the customers in the recycling scheme due to uncertainty of the kitchen waste.

The fuzzy opportunity constraint planning model constructed based on the fuzzy credibility theory is as follows:

where V ═ 0 ═ vu { (0 })₀，V₀1,2,3 …, n is the set of customer points, 0 represents the yard, n is the total number of customers; k {1,2, … m } is the set of available vehicles in the yard, m is the total number of vehicles; r is^kThe total number of vehicles required for the dispatch protocol; c is the fixed cost of vehicle departure; c. C₁Is the unit travel cost of the vehicle; c. C₂A waiting cost for early arrival of the vehicle; d_ijDistance from client i to client j;is the waiting time of vehicle k at customer i; q is the maximum load capacity of the vehicle;representing the amount of fuzzy kitchen waste of customer j, representing triangular fuzzy numbersA measure of credibility of Cr ∈ [0,1 ]]The larger Cr is, the greater credibility that the remaining load capacity of the vehicle can meet the fuzzy demand of the client is; α is a given preference value; b_0kIs the departure time of vehicle k; a is_0kTime for vehicle k to return to yard; e.g. of the type₀An open time window for the yard; l₀The closing time of the yard; t is t_ijA travel time for the vehicle from customer i to customer j;time of arrival of vehicle k at customer i;time of departure for vehicle k from customer i; s_iA service time for the vehicle at customer i; e.g. of the type_iThe start time of the time window is client i; l_iThe closing time of the time window is client i; x is the number of_ijkAs a decision variable, if vehicle k is 1 from customer i to customer j, otherwise it is 0; y is_ikFor decision variables, 1 if customer i is served by vehicle k, and 0 otherwise.

The optimized scheduling method based on the discrete artificial bee colony algorithm comprises the following specific steps of:

step2.1, encoding and decoding mode: integer compilation using customer-based permutationCode policy. The number of the train yard is 0, and the number of the customer is an integer larger than 0. During coding, the customer arrangement after the removal of the parking lot is adopted for coding;the individual i was ranked based on customer for the gen generation. When decoding, according to the capacity of the vehicle and the time window constraint of the client, decoding is carried out by adopting a mode of first routing and then grouping.Is a pair ofA complete delivery scheme is obtained after decoding, S isThe length of S is more than or equal to 2+ n and less than or equal to 2n + 1.

Step2.2, population initialization: and generating SN initial populations with diversity and dispersity by adopting a random rule.

Step2.3, bee hiring phase: the employment bees are placed on the honey sources, one employment bee is assigned to each honey source and the quality of the honey source is evaluated. The number of employed bees is SN. The hiring bee finds a new honey source in the vicinity of the allocated honey source by means of formula (15).

In the formula X_i，X_kRespectively represent the ith and kth honey sources, i ≠ k.Represents a pair X_iAnd X_kPerforming on the basis of r [ a, b]Sequential interleaving of (1), r [ a, b ]]Represents a randomly generated length for controlling the distance of the crossover, where 1 ≦ a ≦ b ≦ n. When the hiring bee finds a new honey source, the new honey source is evaluated, and if the new honey source is X_kThe fitness value of the honey source is superior to that of the old honey source X_iThe fitness of (2) is X_kSubstitution of X_iOtherwise, the old honey source is reserved.

Step2.4, bee stage observation: when all the employed bees complete the search of new honey sources, the observation bees calculate the probability p of each honey source being selected according to the information of the honey sources searched by the employed bees by the formulas (16) and (17), and then the honey sources are randomly selected by adopting the roulette method, wherein p in the formula (16)_iRepresenting the probability that the honey source i is selected, fitness_iThe fitness value of the ith honey source is expressed, and the fitness value is calculated as shown in formula (17), wherein f is shown in formula (17)_iAnd expressing the objective function value of the ith honey source. The chosen honey sources are then searched according to the same search strategy as the employed bees. The number of observation bees is the same as the number of employment bees, both SN.

Step2.5, bee detection stage: after all observation bees finish the search of new honey sources, the selection judgment is carried out on all honey sources, if a certain honey source still cannot be updated after limit cycles, the food source X is_iWill be discarded and the hiring bee corresponding to the food source will be transformed into a scout bee. The scout bee is replaced by a new food source generated by equation (18).

X_i'＝Swap(X_i,u,v) (18)

Equation (18) is for X_iPerforming a swap neighborhood operation, randomly in sequence X_iTwo positions u and v are picked and then the customers in the u and v positions are interchanged.

Step2.6, stop conditions: and setting the termination condition as the maximum iteration number, outputting an optimal path if the maximum iteration number is met, and otherwise, repeating Step2.3, Step2.4 and Step2.5 until the termination condition is met.

The rescheduling method comprises the following specific steps:

and Step3.1, acquiring the recovery path of the current vehicle.

And Step3.2, judging whether the vehicle has the remaining space, if so, turning to Step3.3, otherwise, returning to the parking lot, and rescheduling the remaining unserviced customers by adopting a discrete artificial bee colony algorithm.

Step3.3, whether the service of the customer distributed by the current vehicle is finished or not, if not, the service is continuously carried out according to the distributed path, and if the service is finished, the algorithm is ended.

The invention has the beneficial effects that: the method optimizes the optimization target by determining a vehicle scheduling model and the optimization target of the kitchen waste in the recycling process and designing an optimization scheduling method of a discrete artificial bee colony algorithm and a rescheduling method. The dispatching method is reasonable and effective, and can obtain a good solution of the vehicle optimization dispatching problem in the kitchen waste recovery process in a short time, so that the recovery process of the urban kitchen waste becomes clearer and more accurate, the efficiency of kitchen waste recovery can be greatly improved, and the cost generated in the recovery process is reduced.

Drawings

FIG. 1 is a schematic diagram of vehicle optimized dispatching in a recycling process of kitchen waste according to the present invention;

FIG. 2 is a schematic diagram of a sequential crossover operation of the present invention;

FIG. 3 is a schematic diagram of the operation of the switching neighborhood of the present invention;

fig. 4 is an overall algorithm flow diagram of the present invention.

In fig. 1 to 3, the numeral 0 represents a yard, and the other numerals represent customer numbers.

Detailed Description

Example 1: as shown in fig. 1 to 4, a vehicle optimization scheduling method for a kitchen waste recycling process includes the following steps:

step1, constructing a fuzzy opportunity constraint planning model of the vehicle optimization scheduling problem in the kitchen waste recovery process based on a fuzzy credibility theory, and minimizing the total economic cost summarized by the departure cost, the travel path cost and the time cost of the vehicle as an optimization target.

And Step2, solving the vehicle optimized dispatching problem determined in Step1 in the kitchen waste recovery process by adopting an optimized dispatching method of a discrete artificial bee colony algorithm, and obtaining a recovery scheme before the kitchen waste recovery of the vehicle.

And Step3, recycling the kitchen waste for the customers according to the recycling scheme in Step2, and adjusting the recycling scheme in real time by adopting a rescheduling method so as to reduce the extra cost caused by service failure of part of the customers in the recycling scheme due to uncertainty of the kitchen waste.

Further, the fuzzy opportunity constraint planning model constructed based on the fuzzy credibility theory is as follows:

where V ═ 0 ═ vu { (0 })₀，V₀1,2,3 …, n is the set of customer points, 0 represents the yard, n is the total number of customers; k {1,2, … m } is the set of available vehicles in the yard, m is the total number of vehicles; r is^kThe total number of vehicles required for the dispatch protocol; c is the fixed cost of vehicle departure; c. C₁Is the unit travel cost of the vehicle; c. C₂A waiting cost for early arrival of the vehicle; d_ijDistance from client i to client j;is the waiting time of vehicle k at customer i; q is the maximum load capacity of the vehicle;representing the amount of fuzzy kitchen waste of customer j, representing triangular fuzzy numbersA measure of credibility of Cr ∈ [0,1 ]]The larger Cr is, the greater credibility that the remaining load capacity of the vehicle can meet the fuzzy demand of the client is; α is a given preference value; b_0kIs the departure time of vehicle k; a is_0kTime for vehicle k to return to yard; e.g. of the type₀An open time window for the yard; l₀The closing time of the yard; t is t_ijA travel time for the vehicle from customer i to customer j;time of arrival of vehicle k at customer i;time of departure for vehicle k from customer i; s_iA service time for the vehicle at customer i; e.g. of the type_iThe start time of the time window is client i; l_iThe closing time of the time window is client i; x is the number of_ijkAs a decision variable, if vehicle k is 1 from customer i to customer j, otherwise it is 0; y is_ikFor decision variables, 1 if customer i is served by vehicle k, and 0 otherwise.

Further, the optimized scheduling method based on the discrete artificial bee colony algorithm may be set as follows:

step2.1, encoding and decoding mode: and adopting an integer coding and decoding strategy based on client arrangement. The number of the train yard is 0, and the number of the customer is an integer larger than 0. During coding, the customer arrangement after the removal of the parking lot is adopted for coding;the individual i was ranked based on customer for the gen generation. When decoding, according to the capacity of vehicle and time window constraint of customerAnd decoding the packet after routing.Is a pair ofA complete delivery scheme is obtained after decoding, S isThe length of S is more than or equal to 2+ n and less than or equal to 2n + 1.

Step2.2, population initialization: and generating SN initial populations with diversity and dispersity by adopting a random rule.

Step2.3, bee hiring phase: the employment bees are placed on the honey sources, one employment bee is assigned to each honey source and the quality of the honey source is evaluated. The number of employed bees is SN. The hiring bee finds a new honey source in the vicinity of the allocated honey source by means of formula (15).

In the formula X_i，X_kRespectively represent the ith and kth honey sources, i ≠ k.Represents a pair X_iAnd X_kPerforming on the basis of r [ a, b]Sequential interleaving of (1), r [ a, b ]]Represents a randomly generated length for controlling the distance of the crossover, where 1 ≦ a ≦ b ≦ n. When the hiring bee finds a new honey source, the new honey source is evaluated, and if the new honey source is X_kThe fitness value of the honey source is superior to that of the old honey source X_iThe fitness of (2) is X_kSubstitution of X_iOtherwise, the old honey source is reserved.

Step2.4, bee stage observation: when all the employed bees complete the search of new honey sources, the observation bees calculate the probability p of each honey source being selected according to the information of the honey sources searched by the employed bees by the formulas (16) and (17), and then the honey sources are randomly selected by adopting the roulette method, wherein p in the formula (16)_iRepresenting a source of honeyi probability of being selected, fitness_iThe fitness value of the ith honey source is expressed, and the fitness value is calculated as shown in formula (17), wherein f is shown in formula (17)_iAnd expressing the objective function value of the ith honey source. The chosen honey sources are then searched according to the same search strategy as the employed bees. The number of observation bees is the same as the number of employment bees, both SN.

Step2.5, bee detection stage: after all observation bees finish the search of new honey sources, the selection judgment is carried out on all honey sources, if a certain honey source still cannot be updated after limit cycles, the food source X is_iWill be discarded and the hiring bee corresponding to the food source will be transformed into a scout bee. The scout bee is replaced by a new food source generated by equation (18).

X_i'＝Swap(X_i,u,v) (18)

Equation (18) is for X_iPerforming a swap neighborhood operation, randomly in sequence X_iTwo positions u and v are picked and then the customers in the u and v positions are interchanged.

Step2.6, stop conditions: and setting the termination condition as the maximum iteration number, outputting an optimal path if the maximum iteration number is met, and otherwise, repeating Step2.3, Step2.4 and Step2.5 until the termination condition is met.

Further, the rescheduling method comprises the following specific steps:

and Step3.1, acquiring the recovery path of the current vehicle.

And Step3.2, judging whether the vehicle has the remaining space, if so, turning to Step3.3, otherwise, returning to the parking lot, and rescheduling the remaining unserviced customers by adopting a discrete artificial bee colony algorithm.

Step3.3, whether the service of the customer distributed by the current vehicle is finished or not, if not, the service is continuously carried out according to the distributed path, and if the service is finished, the algorithm is ended.

Specifically, the following can be set: the population size SN in step2.2 is set to 30, and the number of non-updates limit is 5.

The operation results of the vehicle optimized scheduling problem in the process of solving kitchen waste recovery with different scales (different numbers of customers) based on the optimized scheduling scheme of the discrete artificial bee colony algorithm and the standard artificial bee colony algorithm are shown in the table 1. Each algorithm is independently repeated for 20 times on each example, and MAX, MIN and AVG respectively represent the maximum value, the minimum value and the average value of 20 times.

TABLE 1 comparison of objective function values obtained by two algorithms for different problem scales

As can be seen from Table 1, the MAX, MIN and AVG of the discrete artificial bee colony algorithm provided by the invention are superior to those of the original artificial bee colony algorithm in different problem scales, which shows that the optimal scheduling scheme based on the discrete artificial bee colony algorithm can quickly and effectively solve the vehicle optimal scheduling problem in the kitchen waste recovery process.

While the present invention has been described in detail with reference to the embodiments shown in the drawings, the present invention is not limited to the embodiments, and various changes can be made without departing from the spirit of the present invention within the knowledge of those skilled in the art.