CN114611864A - Garbage vehicle low-carbon scheduling method and system - Google Patents

Garbage vehicle low-carbon scheduling method and system Download PDF

Info

Publication number
CN114611864A
CN114611864A CN202111508631.0A CN202111508631A CN114611864A CN 114611864 A CN114611864 A CN 114611864A CN 202111508631 A CN202111508631 A CN 202111508631A CN 114611864 A CN114611864 A CN 114611864A
Authority
CN
China
Prior art keywords
garbage
vehicle
individual
new
point
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202111508631.0A
Other languages
Chinese (zh)
Inventor
申晓宁
潘红丽
陈庆洲
葛忠佩
徐继勇
姚铖滨
许笛
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Nanjing University of Information Science and Technology
Original Assignee
Nanjing University of Information Science and Technology
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Nanjing University of Information Science and Technology filed Critical Nanjing University of Information Science and Technology
Priority to CN202111508631.0A priority Critical patent/CN114611864A/en
Publication of CN114611864A publication Critical patent/CN114611864A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
    • G06Q10/06312—Adjustment or analysis of established resource schedule, e.g. resource or task levelling, or dynamic rescheduling
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/004—Artificial life, i.e. computing arrangements simulating life
    • G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
    • G06Q10/06313—Resource planning in a project environment

Abstract

The invention discloses a garbage vehicle low-carbon scheduling method and a system, comprising the following steps: acquiring input information, including: number of refuse points to be serviced by refuse vehicleNCoordinate information of garbage throwing points, parking lot coordinate information, garbage transfer station coordinate information, garbage amount of each garbage throwing point and garbage vehicle capacityQAnd maximum operating time of the driverT max (ii) a And inputting the input information into a pre-constructed garbage vehicle low-carbon scheduling optimization model based on the adaptive particle swarm optimization algorithm to determine an optimal scheduling scheme. The advantages are that: the method comprises the steps of establishing a garbage clearing multi-stroke low-carbon vehicle scheduling problem model containing practical factors such as vehicle capacity, low carbon, driver working time, multi-stroke and the like, wherein the model is characterized in that a vehicle is allowed to have multiple garbage transfer stations and garbage throwing stationsThe individual trip, in addition, takes into account the environmental pollution problem caused by the vehicle during driving, converts the carbon emission into carbon emission cost and accounts for the total cost.

Description

Garbage vehicle low-carbon scheduling method and system
Technical Field
The invention relates to a low-carbon dispatching method and system for garbage vehicles, and belongs to the technical field of vehicle dispatching.
Background
The global population growth and urbanization are accelerating, urban population is gradually concentrated, living standard and environment are greatly changed, and the more domestic garbage is accumulated. 21/1/2020, the estimated data released by the institute of waste engineering, tokyo, japan, shows that the amount of garbage generated in one year worldwide will reach 320 hundred million tons in 2050, which is 4.2 times that in 2000. The accumulation of garbage causes rapid deterioration of global environment, and the emission of CO2 is in a straight-line trend. According to the data of the International Energy Agency (IEA), the increase of global carbon emissions in 2019 is about 330 hundred million tons. In the face of such tremendous pressures and challenges, high global emphasis has been placed on solid waste integrated management (ISWM) by many local governments. They advocate to develop an integrated municipal solid waste management system. The system comprises front-end garbage generation and terminal garbage disposal, and garbage transportation is connected between the front-end garbage generation and the terminal garbage disposal, so that the transportation link plays a key role in the urban household garbage management system. Among the costs of waste disposal, the cost of transportation is a significant proportion, as the literature indicates that the total annual cost of waste disposal in the united states is about $ 200 billion, with the cost of transportation exceeding $ 100 billion. In addition, transportation links are also one of the main causes of global increases in carbon emissions. In summary, there is a need to develop a deep research on the dispatching of garbage transportation vehicles to reduce transportation cost and environmental pollution.
The vehicle scheduling problem belongs to an NP problem, and as the scale of the problem increases, the accurate algorithm cannot obtain a solution meeting the requirement in the effective time, so that scholars provide a group intelligent optimization algorithm. The particle swarm algorithm (PSO) proposed by Kennedy and Eberhart in 1995 is a representative of these. The self-adaptive particle swarm optimization is an improved version of the particle swarm optimization, integrates problem information of garbage vehicle transportation with the structural characteristics of the particle swarm optimization, and basically comprises the following steps: randomly generating an initial population by adopting an integer coding mode, sequentially decoding individuals according to a decoding mode of eliminating time and capacity constraints, calculating the fitness value of each individual, and determining the individual extreme value and the global extreme value of each individual; carrying out fine search on the decoded individuals by adopting an enhanced local search strategy, selecting a 2-opt algorithm to carry out local mining around the individuals after population decoding according to the capacity and distance information of the problem, and operating only the journey with the largest number of garbage throwing points contained in each individual in order to prevent population diversity loss caused by excessive local search; by adopting a self-adaptive learning strategy based on the contribution degree, from 4 purposes of quickly converging to a global optimal solution, increasing population diversity, exploring a new area and developing an individual optimal position, 4 learning strategies with different advantages are designed, and the most suitable learning strategy is selected for each particle in a self-adaptive manner based on the difference of the contribution degree to generate child particles. The characteristic that the garbage vehicle low-carbon vehicle dispatching planning is solved by the traditional particle swarm algorithm is not utilized, the search is extremely blind, the convergence speed is low, the traditional particle swarm algorithm has the defects of easiness in falling into local optimization, low solving precision and the like, and in conclusion, the vehicle dispatching planning method which is high in convergence efficiency and has strong capability of jumping out of the local optimization is very necessary.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provide a garbage vehicle low-carbon scheduling method and system, which can greatly improve the convergence speed of an algorithm and have strong capability of jumping out of local optimum, thereby quickly planning a group of vehicle scheduling schemes with minimum carbon emission and total transportation cost.
In order to solve the technical problem, the invention provides a low-carbon dispatching method for garbage vehicles, which comprises the following steps:
acquiring input information, including: the number N of garbage throwing points needing service of garbage vehicles and garbageThe coordinate information of the garbage throwing points, the coordinate information of the parking lot, the coordinate information of the garbage transfer station, the garbage amount of each garbage throwing point, the garbage vehicle capacity Q and the maximum working time T of the drivermax；
Inputting input information into a pre-constructed garbage vehicle low-carbon scheduling optimization model based on a self-adaptive particle swarm algorithm, and determining an optimal scheduling scheme;
the optimization goal of the garbage vehicle low-carbon scheduling optimization model based on the adaptive particle swarm optimization is that the total transportation cost of vehicles in a planned scheduling scheme is the minimum, the constraint conditions of the garbage vehicle low-carbon scheduling optimization model based on the adaptive particle swarm optimization are that all vehicles start from a vehicle yard and only start once, each garbage throwing point only allows one vehicle to be served once, when each garbage throwing point is served, one vehicle must run from a certain place to the garbage throwing point and leave from the point, all vehicles completely empty garbage at a garbage transfer station, the garbage loading capacity of the vehicles in one stroke is not greater than the capacity limit of the vehicles, and the working time of a driver of each vehicle cannot exceed the specified maximum working time limit.
Further, the step of inputting the input information into a pre-constructed garbage vehicle low-carbon scheduling optimization model based on the adaptive particle swarm optimization algorithm to determine an optimal scheduling scheme comprises the following steps:
step 1, setting the evolution population scale of an adaptive particle swarm algorithm as n and the maximum neighborhood search scale Y; the maximum evaluation frequency is EvamaxThe evaluation number counter Eva is 0;
step 2, randomly generating N individuals by adopting integer codes, wherein the code of each individual is a string of sequences consisting of integers from 3 to N:
X＝{x3,x4,xi,…,xN}
wherein x isiA number indicating a garbage throwing point, i is 3,4, …, N, a yard number is 1, a garbage transfer station number is 2, and garbage throwing points are 3 to N;
calculating a target value f (x) for each individual:
f(X)＝Cfixed+Cfuel+Ccarbon
wherein, CfixedTo fix the cost, CfuelFor fuel cost, CcarbonIs the carbon emission cost;
Figure BDA0003404383290000031
wherein, CfFor a fixed use cost per vehicle, UkIndicating whether the K-th vehicle is used, K indicating the total number of vehicles,
Figure BDA0003404383290000032
Figure BDA0003404383290000033
wherein, CmThe fuel cost per unit distance traveled by a vehicle, B represents the total travel set for a vehicle, dijThe distance between the garbage throwing point i and the garbage throwing point j,
Figure BDA0003404383290000034
indicating whether the kth vehicle has traveled the path (i, j) on the b-th trip,
Figure BDA0003404383290000035
ECij＝FE·FCij
Figure BDA0003404383290000036
wherein, ECijRepresenting the carbon emission of a vehicle travelling from a refuse dump point i to j, FE being a fuel emission parameter, CeIn order to be a carbon tax, the carbon tax is,
Figure BDA0003404383290000037
indicating whether the kth vehicle is servicing a point of trash placement i on the b-th trip,
Figure BDA0003404383290000041
FCijthe specific calculation mode is that the fuel oil consumed by the vehicle from the garbage throwing point i to the garbage throwing point j is as follows:
FCij＝[αij(z+lij)+βv2]dij
wherein z represents the vehicle weight,/ijRepresents the load of the vehicle traveling from the refuse deposit point i to j, v represents the traveling speed of the vehicle, and αijAnd β are parameters related to road conditions and vehicle types, respectively, and the calculation method is as follows:
αij＝a+gsinθij+gCrcosθij
β＝0.5CdAρ
where a is the vehicle running acceleration, g is the gravitational acceleration constant, and θijRoad surface gradient, C, for the section from refuse deposit point i to refuse deposit point jrIs a coefficient of rolling resistance, CdIs the traction coefficient, A is the vehicle frontal surface area, ρ represents the air density;
fitness of the individual is f (x):
Figure BDA0003404383290000042
step 3, finely searching the decoded individuals by adopting an enhanced local search strategy to form new individuals;
step 4, adopting a self-adaptive learning strategy based on the contribution degree for the new individual, and self-adaptively selecting a learning mode most conforming to the self stage of the particle to generate a child particle:
step 5, updating the individual extreme value and the global extreme value in each iteration according to the rule of the winner and the disadvantage;
step 6, if Eva>EvamaxAnd (3) terminating iteration, outputting the individual with the optimal fitness, wherein the individual is the planned vehicle scheduling scheme, and otherwise, correspondingly increasing the Eva and turning to the step 3.
Further, the performing a fine search on the decoded individuals by using the enhanced local search strategy to form new individuals includes:
step 31, decoding the individuals in a decoding mode of eliminating time and capacity constraints to obtain a dispatching scheme of the garbage vehicles;
step 32, finding out a vehicle index and a journey index corresponding to the journey with the largest number of garbage throwing points;
step 33, performing 2-opt optimization on the journey containing the largest number of garbage throwing points, and effectively opening a cross route in the journey;
step 34, splicing the optimized journey and other unselected journeys again according to the sequence to form a new individual; during the splicing process, two cases are divided according to the end point of the vehicle travel: when the terminal point of the journey is a transfer station, the vehicle is indicated not to go to the parking lot any more in the journey, and at the moment, only the starting point and the garbage transfer station in the journey need to be removed; when the end point of the travel is the yard, the vehicle goes to a garbage transfer station to unload and then returns to the yard, and at the moment, the departure point in the travel and the yard of the garbage transfer station need to be removed;
and step 35, outputting the new individuals subjected to the enhanced local search strategy.
Further, the generating of the child particle by using a contribution-based adaptive learning strategy for the new individual and adaptively selecting a learning mode that best meets the self stage of the particle includes:
step 41, setting m learning strategies;
step 42, making contribution C of each learning strategykk＝0，kk＝1,2,…,m；
Step 43, selecting and determining a learning strategy, namely, learn (ii), for each new individual ii in the population through roulette respectively, generating new particles, namely, Npop (ii), according to the learn (ii), and calculating a target value of the Npop (ii);
step 44, sorting new particles in the new population according to the target value to obtain a ranking vector r (ii) of the particles;
step 45, assign a weight w to the ii new particle according to the following equationiiThe more the particle rank is, the larger the weight is distributed;
Figure BDA0003404383290000051
wherein, wiiR (ii) is the weight that the ii particle should be assigned, r (ii) is the rank of the ii particle;
and step 46, updating the contribution degree of the kth learning strategy according to the following formula,
Figure BDA0003404383290000061
wherein sfiiIf the fitness value of the ii th particle is improved through the kth learning strategy for the reward factor, sf is carried outii1, otherwise sfii＝0；
Step 47, the method for updating and normalizing the kth learning strategy selection probability is shown as the following formula
pkk＝Ckk+ε kk＝1,2,...,m
Figure BDA0003404383290000062
Wherein p iskkSelecting probabilities for a kth learning strategy, jj representing subscripts of the learning strategies, wherein jj is 1, 2.
And 48, after the particles in the population are learned by selecting different learning modes, outputting all newly generated child particles Npop, the target value Npopbj of the new particle, the contribution degree vector C of the learning strategy and the selection probability vector P.
Further, the m learning strategies include: greedy learning strategies, multivariate learning strategies, exploratory learning strategies and exploratory learning strategies,
the greedy learning strategy is: generating a variation individual by reversing variation of the new individual, generating a cross individual I by the variation individual and an individual extreme value through a greedy cross operator, and performing greedy cross on the cross individual I and a global extreme value to generate a new individual after greedy learning;
the multivariate learning strategy is as follows: generating variant individuals by a new individual through a multivariate variant operator, carrying out partial mapping and crossing on the generated variant individuals and individual extreme values to obtain a cross individual I, and carrying out partial mapping and crossing on the cross individual 1 and a global extreme value to generate a new individual after multivariate learning;
the exploration type learning strategy is that new individuals generate variant individuals through a multivariate variant operator, and partial mapping and crossing are carried out on the generated variant individuals and the individual extreme value of any particle with fitness superior to the population, so as to generate new individuals after exploration and learning;
the utilization type learning strategy is that new individuals generate variant individuals through a multivariate mutation operator, and the generated variant individuals and individual extreme values are subjected to partial mapping and crossing to generate new individuals after utilization learning.
Further, the greedy learning strategy includes:
step 411, determining the individuals X and the individual extremum X which need to be crossedpbestAnd global extremum Xgbest；
Step 412, randomly selecting one point from the points which are not the parking lot and the non-transfer station as a first garbage throwing point S of the vehicle service, and selecting the first garbage throwing point S at XpbestLeft garbage throwing point SLpAnd a right garbage throwing point SRpIn X, the left garbage throwing point SLXAnd a right garbage throwing point SRXAs a candidate set for the next access;
step 413, in candidate set { SLp,SRp,SLX,SRXSelecting a garbage throwing point S 'from the previous step, so that the S' and the S form a path section with the least transportation cost;
step 414, if S' ∈ { S ∈ [ ]Lp,SLXExecuting step 415, otherwise executing step 416;
step 415, from the individual extremum XpbestAnd deleting S from the individual X, setting S' as the next starting point S, and setting S as the left candidate set { SLp,SLXSelecting new S 'as a next garbage putting point for service, enabling S' and S to form a path section with the least transportation cost, and repeatedly executing S515 until a new solution X is generatednew；
Step 416, from the individual extremum XpbestAnd deleting S from the individual X, setting S' as the next starting point S, and setting S as the right candidate set { SRp,SRXSelecting new S 'as a next garbage putting point for service, enabling S' and S to form a path section with the least transportation cost, and repeatedly executing S516 until a new solution X is generatednew；
Step 417, with X and XpbestGreedy intersection of (2) as an example, the new solution XnewAnd global extreme value XgbestPerforming greedy intersection again to generate a new solution;
step 418, output new individuals through greedy crossover operators.
Further, the multivariate learning strategy comprises:
step 421, calculating the selection probability of the reverse variation, the crossover variation and the insertion variation;
step 422, selecting the serial number of the variation mode based on the roulette mode;
step 423, if the sequence number of the reverse variation is selected, the individual is subjected to exchange variation;
step 424, if the sequence number of the crossover variation is selected, the individual is subjected to reverse variation;
step 425, if inserting the serial number of the mutation, inserting the mutation into the individual;
and 426, outputting the new individuals passing through the multivariate mutation operator according to the mutation result.
A low-carbon dispatching system for garbage vehicles comprises:
the acquisition module is used for acquiring input information and comprises: the number N of the garbage throwing points which need to be served by the garbage vehicle, the coordinate information of the garbage throwing points, the coordinate information of the parking lot, the coordinate information of the garbage transfer station, and the garbage of each garbage throwing pointVolume, garbage vehicle capacity Q and maximum driver operating time Tmax；
The determining module is used for inputting input information into a pre-constructed garbage vehicle low-carbon scheduling optimization model based on the adaptive particle swarm optimization algorithm and determining an optimal scheduling scheme;
the optimization goal of the garbage vehicle low-carbon scheduling optimization model based on the adaptive particle swarm optimization is that the total transportation cost of vehicles in a planned scheduling scheme is the minimum, the constraint conditions of the garbage vehicle low-carbon scheduling optimization model based on the adaptive particle swarm optimization are that all vehicles start from a vehicle yard and only start once, each garbage throwing point only allows one vehicle to be served once, when each garbage throwing point is served, one vehicle must run from a certain place to the garbage throwing point and leave from the point, all vehicles completely empty garbage at a garbage transfer station, the garbage loading capacity of the vehicles in one stroke is not greater than the capacity limit of the vehicles, and the working time of a driver of each vehicle cannot exceed the specified maximum working time limit.
A computer readable storage medium storing one or more programs, the one or more programs comprising instructions, which when executed by a computing device, cause the computing device to perform any of the methods.
A computing device, comprising:
one or more processors, memory, and one or more programs stored in the memory and configured to be executed by the one or more processors, the one or more programs including instructions for performing any of the methods.
The invention achieves the following beneficial effects:
(1) and establishing a garbage clearing multi-stroke low-carbon vehicle dispatching problem model containing practical factors such as vehicle capacity, low carbon, driver working time, multi-stroke and the like. The model features multiple trips that a vehicle can take between the yard, the terminal and the terminal. In addition, considering the environmental pollution problem caused during the driving of the vehicle, the carbon emission amount is converted into a carbon emission cost and is counted into the total cost.
(2) The invention adopts a garbage vehicle low-carbon scheduling method based on the self-adaptive particle swarm algorithm, eliminates the decoding mode of time and capacity constraint, and ensures that the decoded solutions are all feasible solutions, so the performance of the algorithm is superior to that of the traditional particle swarm algorithm.
(3) The enhanced local search strategy is used for locally searching the most strokes containing the garbage throwing points in the individuals, so that the phenomenon of over-fast population assimilation caused by excessive local search is avoided while the algorithm solving precision is improved.
(4) Based on the self-adaptive learning mechanism of the contribution degree, four learning strategies with different advantages are designed for individuals in different evolution stages: greedy learning, multivariate learning, exploratory learning and utilization learning, and appropriate learning strategies are self-adaptively distributed to individuals according to different contribution degrees of the greedy learning, the multivariate learning, the exploratory learning and the utilization learning in the population evolution process, so that the search efficiency of the algorithm is improved.
Drawings
FIG. 1 is a main flow chart of the present invention using adaptive particle swarm optimization;
FIG. 2 is a diagram illustrating an optimal solution obtained by solving an example using the adaptive particle swarm optimization of the present invention;
FIG. 3 is a graph of the results of comparing the fixed cost, fuel cost and carbon emission cost of the optimal solution using the adaptive particle swarm optimization of the present invention with the existing algorithm for solving the garbage disposal problem.
Detailed Description
The invention is further described below with reference to the accompanying drawings. The following examples are only for illustrating the technical solutions of the present invention more clearly, and the protection scope of the present invention is not limited thereby.
Through on-site investigation of a green ring company in a new area of JiangBei, Nanjing, China, 54 cells (marked as 3-56) longitude and latitude coordinates served by the company, a garbage transfer station (marked as 2) and a parking lot (marked as 1) longitude and latitude coordinates, and the garbage amount of each cell in a certain day are obtained. The longitude and latitude coordinates of all points are converted into the coordinates of a plane rectangular coordinate system as shown in table 1. The capacity of the garbage vehicle is limited to 10t, and the maximum duration of the driver is limited to 8 h.
TABLE 1
Figure BDA0003404383290000091
Figure BDA0003404383290000101
Figure BDA0003404383290000111
The optimal vehicle dispatching scheme obtained by solving the embodiment based on the adaptive particle swarm algorithm provided by the invention is shown in a main flow chart of fig. 1, and comprises the following specific steps:
and S1, initializing. Reading the input information of the example (see table 1); and (4) giving the definition of an optimization target and setting a constraint condition.
Plane coordinate information of garbage throwing point (A)x1,Ay1),(Ax2,Ay2),…,(AxN,AyN) And the scale of the problem indicates that the problem contains N points (the number of the parking lot is 1, the number of the garbage transfer station is 2, and the number of the garbage throwing points is 3-N), and the distance between different points is an Euclidean distance calculation formula which is defined as:
Figure BDA0003404383290000112
wherein d isijRepresenting the distance between a garbage throwing point i and a garbage throwing point j;
the optimization objective "transportation cost incurred in route" is defined as:
f(X)＝Cfixed+Cfuel+Ccarbon
wherein f (X) includes a fixed cost CfixedFuel cost CfuelAnd carbon emission cost Ccarbon。
(a) Fixing cost:
once the vehicle is used during the garbage clearing process, the corresponding maintenance cost and the compensation cost of the driver are generated. Therefore, the fixed cost C generated by completing the disposal of the garbage oncefixedThe following were used:
Figure BDA0003404383290000113
(b) fuel cost:
in the driving process of the vehicle, fuel cost is generated due to fuel consumption, and the fuel consumption efficiency is often influenced by factors such as driving speed and road conditions. The vehicle is assumed to have stable road conditions and uniform speed during transportation, and the fuel cost generated by the vehicle per unit distance is fixed. Therefore, all fuel costs C incurred after completing a full trip using the vehiclefuelThe following were used:
Figure BDA0003404383290000121
(c) carbon emission cost:
due to the aggravation of greenhouse effect, many countries have developed carbon tax regulations to control CO2And (4) discharging. Carbon tax refers to the treatment of CO2And charging corresponding fees according to the discharge amount. Therefore, all the carbon emission cost C generated after the vehicle is used to complete the entire journeycarbonThe following were used:
ECij＝FE·FCij
Figure BDA0003404383290000122
Figure BDA0003404383290000123
Figure BDA0003404383290000124
Figure BDA0003404383290000125
wherein FE is a fuel oil discharge parameter, CeAs a carbon tax, CmCost of fuel spent for a unit distance traveled by the vehicle, CfFor a fixed cost per vehicle, B represents the total travel set for a vehicle, dijIs the distance between the refuse drop point i and the refuse drop point j, FCijThe specific calculation mode of the fuel oil consumed by the vehicle driving from the garbage throwing point i to the garbage throwing point j is as follows:
FCij＝[αij(z+lij)+βv2]dij
wherein alpha isijAnd β are parameters related to road conditions and vehicle types, respectively, and the calculation method is as follows:
αij＝a+gsinθij+gCrcosθij
β＝0.5CdAρ
where a is the vehicle running acceleration, g is the gravitational acceleration constant, and θijRoad surface gradient, C, for the section from refuse deposit point i to refuse deposit point jrIs a coefficient of rolling resistance, CdFor the traction coefficient, A is the vehicle frontal surface area and ρ represents the air density.
Defining constraints includes the following six:
(1) ensure that all vehicles start from the yard only once, i.e.:
Figure BDA0003404383290000131
Figure BDA0003404383290000132
(2) each trash can point allows one vehicle to service only once, namely:
Figure BDA0003404383290000133
(3) when each trash can is served, a vehicle must travel from a location to the trash can and leave from the location, namely:
Figure BDA0003404383290000134
Figure BDA0003404383290000135
(4) all vehicles empty the garbage completely at the garbage transfer station, namely:
Figure BDA0003404383290000136
(5) the load of the vehicle in one stroke is not larger than the limit of the capacity of the vehicle, namely:
Figure BDA0003404383290000137
(6) the operating time of each vehicle must not exceed a specified maximum operating time limit, namely:
Figure BDA0003404383290000138
wherein q isiIndicates the amount of refuse, l, at each refuse drop point iijRepresenting the load of the vehicle travelling from refuse deposit point i to j, tijRepresenting the time, T, for the vehicle to travel from refuse drop point i to jmaxIndicating the maximum daily operating time of the driver.
S2, initializing improved particle swarm algorithm parameters based on heuristic information:
is provided withThe self-adaptive particle swarm algorithm has the evolution population size of n being 200, the neighborhood solution number Y being 10 and the maximum evaluation times being EvamaxThe evaluation count counter Eva is set to 0 at 100000.
S3, generating an initial candidate population, and calculating the fitness:
by adopting integer coding, for the problem of containing N points (the number of a parking lot is 1, the number of a garbage transfer station is 2, and the number of a garbage throwing point is 3-N), the codes of each individual are a string of sequences consisting of integers from 3 to N:
X＝{x3,x4,xi,…,xN}
wherein x isi(i-3, 4, …, N) denotes the number of points where refuse is deposited; according to the known optimization objective in the step (1), the transportation cost generated in the path is taken as the optimization objective, that is, the less the transportation cost is, the higher the fitness is, and the better the planned scheme is, the individual fitness is defined as:
Figure BDA0003404383290000141
an individual extremum and a global extremum of the individual are determined.
S4, fine search is carried out on the decoded individuals by adopting the enhanced local search strategy:
s41, decoding the individuals in a decoding mode of eliminating time and capacity constraints to obtain a dispatching scheme of the garbage vehicle;
s42, finding out the vehicle index and the journey index corresponding to the journey containing the largest number of garbage throwing points;
s43, performing 2-opt optimization on the journey containing the largest number of garbage throwing points, and effectively opening a cross route in the journey;
s44, splicing the optimized journey and the rest unselected journeys again according to the sequence to form a new individual;
s45, during the splicing process, the two cases are divided according to the end point of the vehicle travel: when the end point of the journey is a transfer station, the vehicle is indicated not to go to the yard any more in the journey. Therefore, only the start point and the end point (garbage transfer station) in the trip need to be removed. When the end point of the journey is a yard, the vehicle must go to a garbage transfer station to unload and then returns to the yard, so that a starting point, a garbage transfer station and an end point (the yard) in the journey need to be removed;
and S46, outputting the new individual subjected to the enhanced local search strategy.
S5, adopting a self-adaptive learning strategy based on contribution degree, self-adaptively selecting a learning mode most conforming to the self stage of the particle, and generating the child particle comprises the following implementation steps:
s51, setting a greedy learning strategy:
the individual generates variant individuals by reversing the variation. And secondly, generating a crossed individual 1 by the variant individual and the individual extreme value through a greedy cross operator. Finally, carrying out greedy crossing on the crossing individual 1 and the global extreme value in the same way to generate a new individual after greedy learning;
s52, setting a multivariate learning strategy:
the particles generate variant individuals through a multivariate mutation operator. And secondly, carrying out partial mapping and crossing on the generated variant individuals and the individual extreme values to obtain crossed individuals 1. Finally, carrying out partial mapping intersection on the intersection individual 1 and the global extreme value to generate a new individual after multivariate learning;
s53, setting a exploration type learning strategy:
the particles generate variant individuals through a multivariate mutation operator. Secondly, carrying out partial mapping and crossing on the generated variant individuals and the individual extreme value of any particle with fitness superior to the individual extreme value in the population to generate new individuals after exploration and learning;
and S54, setting a utilization learning strategy, and generating variant individuals by the particles through a multivariate mutation operator. Secondly, carrying out partial mapping and crossing on the generated variant individuals and individual extremum values to generate new individuals after learning;
s55, during population initialization, making contribution C of each learning strategykk0, kk 1,2, …, m; (assuming a total of m learning strategies);
s56, selecting and determining a learning strategy, namely, learn (ii), for each new individual ii in the population through roulette respectively, generating new particles Npop (ii) according to the learn (ii) and calculating a target value of the Npop (ii);
s57, sorting new particles in the new population according to the target value to obtain the ranking r (ii) of the particles ii;
s58, assigning a weight w to the ii th new particle according to the following formulaiiThe more the particle is ranked, the greater the weight assigned.
Figure BDA0003404383290000161
Wherein, wiiR (ii) is the weight that the ii particle should be assigned, and r (ii) is the rank of the ii particle.
S59, the contribution degree of the kth learning strategy is updated as follows.
Figure BDA0003404383290000162
Wherein sfiiIf the fitness value of the ii th particle is improved through the kth learning strategy for the reward factor, sf is carried outii1, otherwise sfii＝0。
S510, updating and normalizing the kk learning strategy selection probability method as shown in the following formula
pkk＝Ckk+ε kk＝1,2,...,m
Figure BDA0003404383290000163
Wherein p iskkThe probability is selected for the kth learning strategy, jj denotes the subscript of the learning strategy, and jj 1, 2.
S511, after the particles in the population are learned by selecting different learning manners, all the newly generated child particles Npop, the target value npobj of the new particle, the contribution vector C of the learning strategy, and the selection probability vector P are output.
And S6, updating the individual extremum and the global extremum: and updating the individual extremum and the global extremum in each iteration according to the rule of the winning or the losing.
S7, judging termination criteria: if Eva>EvamaxThe iteration is terminated, and the individuals with the optimal fitness are output, wherein the individuals are the planned sequence for accessing the garbage throwing points, otherwise, the Eva is correspondingly increased, and the step S4 is carried out.
In step S51, the greedy crossover operator in the design greedy learning strategy is specifically implemented as follows:
s511, determining the individual X and the individual extreme value X which need to be crossedpbestAnd global extremum Xgbest；
S512, randomly selecting one point from points without parking lots and transfer stations as a first garbage throwing point S for vehicle service, and selecting the garbage throwing point S at XpbestLeft garbage throwing point SLpAnd a right garbage throwing point SRpIn X, the left garbage throwing point SLXAnd a right garbage throwing point SRXAs a candidate set for the next access;
s513, in the candidate set { SLp,SRp,SLX,SRXSelecting S 'from the S map, so that the S' and the S form a path section with the least transportation cost;
s514, if the client S' is equal to S ∈ { SLp,SLXExecuting S515, otherwise executing S516;
s515, from the individual extremum XpbestAnd deleting S from the individual X, setting S' as the next starting point S, and setting S as the left candidate set { SLp,SLXSelecting a new S 'as a garbage putting point for the next service, so that S' and S form a path segment with the least transportation cost, and repeatedly executing S515 until a new solution X is generatednew；
S516, from the individual extreme value XpbestAnd deleting S from the individual X, setting S' as the next starting point S, and setting S as the right candidate set { SRp,SRXSelecting a new S 'as a next garbage throwing point for service, so that the S' and the S form a path section with the least transportation cost and the weightRepeating S516 until a new solution X is generatednew；
S517, with X and XpbestTake greedy intersection of (a) as an example, solve the new solution XnewAnd global extreme value XgbestPerforming greedy intersection again to generate a new solution;
s518, outputting the new individuals passing through the greedy crossover operator.
As a preferred example, in step S52, the multivariate mutation operator in the design multivariate learning strategy is implemented as follows:
s521, calculating the selection probabilities of reverse variation, crossover variation and insertion variation;
s522, selecting a serial number of a variation mode based on a roulette mode;
s523, if the mutation mode 1 is selected, the individual is subjected to crossover mutation;
s524, if variation mode 2 is selected, performing a reverse variation on the individual;
s525, if the variation mode 3 is selected, performing insertion variation on the individual;
and S526, outputting the new individuals after the multivariate mutation operator passes.
Correspondingly, the invention also provides a low-carbon dispatching system for the garbage truck, which comprises the following components:
the acquisition module is used for acquiring input information and comprises: the number N of garbage throwing points needing to be served by the garbage vehicle, the coordinate information of the garbage throwing points, the coordinate information of a parking lot, the coordinate information of a garbage transfer station, the garbage amount of each garbage throwing point, the capacity Q of the garbage vehicle and the maximum working time T of a drivermax；
The determining module is used for inputting input information into a pre-constructed garbage vehicle low-carbon scheduling optimization model based on the adaptive particle swarm optimization algorithm and determining an optimal scheduling scheme;
the optimization goal of the garbage vehicle low-carbon scheduling optimization model based on the adaptive particle swarm optimization is that the total transportation cost of vehicles in a planned scheduling scheme is the minimum, the constraint conditions of the garbage vehicle low-carbon scheduling optimization model based on the adaptive particle swarm optimization are that all vehicles start from a vehicle yard and only start once, each garbage throwing point only allows one vehicle to be served once, when each garbage throwing point is served, one vehicle must run from a certain place to the garbage throwing point and leave from the point, all vehicles completely empty garbage at a garbage transfer station, the garbage loading capacity of the vehicles in one stroke is not greater than the capacity limit of the vehicles, and the working time of a driver of each vehicle cannot exceed the specified maximum working time limit.
A computer readable storage medium storing one or more programs, the one or more programs comprising instructions, which when executed by a computing device, cause the computing device to perform any of the methods.
A computing device, comprising:
one or more processors, memory, and one or more programs stored in the memory and configured to be executed by the one or more processors, the one or more programs including instructions for performing any of the methods.
The effect of the invention can be further illustrated by the following simulation experiment:
1. the experimental conditions are as follows:
matlab 2017b is used for simulation on an Intel (R) core (TM) i5-5500U CPU @2.40GHz and internal memory 8GB and WINDOWS 10 system.
2. The experimental contents are as follows:
through on-site investigation of a green ring company in a new area of JiangBei, Nanjing, China, 54 cells (marked as 3-56) longitude and latitude coordinates served by the company, a garbage transfer station (marked as 2) and a parking lot (marked as 1) longitude and latitude coordinates, and the garbage amount of each cell in a certain day are obtained. The longitude and latitude coordinates of all points are converted into the coordinates of a plane rectangular coordinate system as shown in table 1. The capacity of the garbage vehicle is limited to 10t, and the maximum duration of the driver is limited to 8 h.
3. Results of the experiment
i. Compared with the existing algorithm for solving the garbage clearing problem, the invention has the advantages that the invention is adopted to compare the experimental results of the garbage clearing example of the green ring company in the new area of Jiangbei of Nanjing;
ii, solving a vehicle dispatching path planning graph of the minimum transportation cost of the garbage clearing problem by adopting the self-adaptive particle swarm optimization;
and iii, comparing the fixed cost, the fuel oil cost and the carbon emission cost by adopting the optimal scheme of the method and the existing algorithm for solving the garbage cleaning problem.
The experiments were run independently 30 times each in the examples. Table 2 lists the comparison results of the comparison algorithm and the adaptive particle swarm algorithm, respectively. Wherein Best and mean represent the optimal and average transportation cost target values searched in 30 runs respectively, and the Best values of Best and mean are represented by bold. The experimental results show that the comprehensive performance of the algorithm on the garbage clearing problem is superior to that of various comparison algorithms, and the solution precision is high. Therefore, the algorithm can effectively solve the garbage clearing problem, and can provide a vehicle scheduling scheme with low transportation cost and less carbon emission for garbage clearing companies.
TABLE 2
Figure BDA0003404383290000201
Fig. 2 is a specific scheduling scheme of the proposed algorithm in an example of scheduling garbage collection vehicles in the new green ring company of the northwest of the jiang, tokyo. The black points represent garbage throwing points, the red five-pointed star represents a garbage yard, and the green square represents a garbage transfer station. The black broken line indicates all the strokes traveled by the vehicle 1, and the red solid line indicates all the strokes traveled by the vehicle 2. As can be seen from fig. 2, a total of two vehicles are required to go out to complete the garbage cleaning task. The first vehicle drives 4 trips under the condition that the maximum time limit of a driver is met, wherein the trips are respectively (1-21-14-28-29-31-32-30-27-22-25-50-2), (2-41-40-37-39-52-46-42-2), (2-4-43-53-34-33-38-36-56-35-2), (2-54-49-55-26-48-47-51-45-44-2-1), the trip arrangement of the second vehicle is (1-13-3-20-10-9-6-5-8-12-2), (2-11-23-15-16-19-17-18-7-24-2-1). Since the optimization goal of this problem is to minimize the transportation cost without having an excessive requirement on the time for completion of the clearing, the garbage clearing task only needs to be completed within the maximum working time of the driver, and thus reducing the use of vehicles is beneficial to saving the transportation cost. In addition, the vehicle 2 completes the collection of the last trash drop point in the second trip, at which time the trash cleaning task has been completed by both vehicles. Thus, the vehicle 2 has only two strokes.
Figure 3 shows the comparison of the proposed algorithm and the optimal solution of the comparison algorithm in terms of fixed costs, fuel costs and carbon emission costs. As can be seen from fig. 3, the TS algorithm has the highest fixed cost because it invokes the most vehicles. The proposed algorithm achieves the lowest carbon emission cost and total transportation cost among all comparison algorithms. Compared to HGA and SA + GA, the proposed algorithm, although slightly higher in fuel cost, has a greater dominance in carbon emissions cost than the disadvantage in fuel cost, and therefore also lower in its total transportation cost. As can be seen from the formula for the formula optimization objective, the fuel cost relates only to the distance information, while the carbon emission cost is related to both the distance and load information. The remaining comparative algorithms in the experiments herein only consider distance information between service points, while the proposed algorithm guides the search operation of the algorithm with both distance and payload as heuristic information, so it is more likely to achieve lower carbon emissions and overall transportation costs. In conclusion, the algorithm is more suitable for solving the garbage cleaning problem, and the transportation cost including the fuel oil cost and the carbon emission cost can be saved to the maximum extent by reasonably arranging vehicles and routes, so that the aim of energy conservation and emission reduction is really achieved.
In this disclosure, aspects of the present invention are described with reference to the accompanying drawings, in which a number of illustrative examples are shown. Examples of the present disclosure are not necessarily defined to include all aspects of the invention. It should be appreciated that the various concepts and examples described above may be embodied in any of numerous ways, as the disclosed concepts and examples are not limited to any embodiment. In addition, some aspects of the present disclosure may be used alone, or in any suitable combination with other aspects of the present disclosure.
Although the present invention has been described with reference to the preferred embodiments, it is not intended to be limited thereto. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the invention. Therefore, the protection scope of the present invention should be determined by the appended claims.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and so forth) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The above description is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, several modifications and variations can be made without departing from the technical principle of the present invention, and these modifications and variations should also be regarded as the protection scope of the present invention.

Claims (10)

1. A low-carbon dispatching method for garbage vehicles is characterized by comprising the following steps:
acquiring input information, including: the number N of garbage throwing points needing to be served by the garbage vehicle, the coordinate information of the garbage throwing points, the coordinate information of a parking lot, the coordinate information of a garbage transfer station, the garbage amount of each garbage throwing point, the capacity Q of the garbage vehicle and the maximum working time T of a drivermax；
Inputting input information into a pre-constructed garbage vehicle low-carbon scheduling optimization model based on a self-adaptive particle swarm algorithm, and determining an optimal scheduling scheme;
the optimization goal of the garbage vehicle low-carbon scheduling optimization model based on the adaptive particle swarm optimization is that the total transportation cost of vehicles in a planned scheduling scheme is the minimum, the constraint conditions of the garbage vehicle low-carbon scheduling optimization model based on the adaptive particle swarm optimization are that all vehicles start from a vehicle yard and only start once, each garbage throwing point only allows one vehicle to be served once, when each garbage throwing point is served, one vehicle must run from a certain place to the garbage throwing point and leave from the point, all vehicles completely empty garbage at a garbage transfer station, the garbage loading capacity of the vehicles in one stroke is not greater than the capacity limit of the vehicles, and the working time of a driver of each vehicle cannot exceed the specified maximum working time limit.
2. The garbage vehicle low-carbon scheduling method of claim 1, wherein the input information is input into a pre-constructed garbage vehicle low-carbon scheduling optimization model based on an adaptive particle swarm algorithm, and an optimal scheduling scheme is determined, and the method comprises the following steps:
step 1, setting the evolution population scale of an adaptive particle swarm algorithm as n and the maximum neighborhood search scale Y; the maximum evaluation frequency is EvamaxThe evaluation frequency counter Eva is 0;
step 2, randomly generating N individuals by adopting integer codes, wherein the code of each individual is a string of sequences consisting of integers from 3 to N:
X＝{x3,x4,xi,…,xN}
wherein x isiA number indicating a garbage throwing point, i is 3,4, …, N, a yard number is 1, a garbage transfer station number is 2, and garbage throwing points are 3 to N;
calculating a target value f (x) for each individual:
f(X)＝Cfixed+Cfuel+Ccarbon
wherein, CfixedTo fix the cost, CfuelFor fuel cost, CcarbonCost for carbon emissions;
Figure FDA0003404383280000021
wherein, CfFor a fixed cost of use for each vehicle, UkIndicating whether the K-th vehicle is used, K indicating the total number of vehicles,
Figure FDA0003404383280000022
Figure FDA0003404383280000023
wherein, CmThe fuel cost per unit distance traveled by a vehicle, B represents the total travel set for a vehicle, dijThe distance between the garbage throwing point i and the garbage throwing point j,
Figure FDA0003404383280000024
indicating whether the k-th vehicle passes through the route in the b-th trip
Figure FDA0003404383280000025
ECij＝FE·FCij
Figure FDA0003404383280000026
Wherein, ECijRepresenting the carbon emission of a vehicle travelling from a refuse dump point i to j, FE being a fuel emission parameter, CeIn order to be a carbon tax, the carbon tax is,
Figure FDA0003404383280000027
indicating whether the kth vehicle is servicing a point of trash placement i on the b-th trip,
Figure FDA0003404383280000028
FCijthe specific calculation mode is that the fuel oil consumed by the vehicle from the garbage throwing point i to the garbage throwing point j is as follows:
FCij＝[αij(z+lij)+βv2]dij
wherein z represents the vehicle weight,/ijRepresenting the load of the vehicle travelling from the refuse deposit point i to j, v representing the travelling speed of the vehicle, alphaijAnd β are parameters related to road conditions and vehicle types, respectively, and the calculation method is as follows:
αij＝a+gsinθij+gCrcosθij
β＝0.5CdAρ
where a is the vehicle running acceleration, g is the gravitational acceleration constant, and θijRoad surface gradient, C, for the section from refuse deposit point i to refuse deposit point jrIs a coefficient of rolling resistance, CdFor traction coefficient, A is the vehicle frontal surface area, ρ chartIndicating the air density;
the fitness of the individual is f (x):
Figure FDA0003404383280000031
step 3, finely searching the decoded individuals by adopting an enhanced local search strategy to form new individuals;
step 4, adopting a self-adaptive learning strategy based on contribution degree for the new individual, and adaptively selecting a learning mode most conforming to the self stage of the particle to generate a filial generation particle:
step 5, updating the individual extreme value and the global extreme value in each iteration according to the rule of the winner and the disadvantage;
step 6, if Eva>EvamaxAnd (3) terminating iteration, outputting the individual with the optimal fitness, wherein the individual is the planned vehicle scheduling scheme, and otherwise, correspondingly increasing the Eva and turning to the step 3.
3. The garbage vehicle low-carbon scheduling method of claim 2, wherein the fine searching of the decoded individuals by using the enhanced local search strategy to form new individuals comprises:
step 31, decoding the individuals in a decoding mode of eliminating time and capacity constraints to obtain a dispatching scheme of the garbage vehicles;
step 32, finding out a vehicle index and a journey index corresponding to the journey with the largest number of garbage throwing points;
step 33, performing 2-opt optimization on the journey containing the largest number of garbage throwing points, and effectively opening a cross route in the journey;
step 34, splicing the optimized journey and other unselected journeys again according to the sequence to form a new individual; during the splicing process, two cases are divided according to the end point of the vehicle travel: when the terminal point of the journey is a transfer station, the vehicle is indicated not to go to the parking lot any more in the journey, and at the moment, only the starting point and the garbage transfer station in the journey need to be removed; when the end point of the travel is the yard, the vehicle goes to a garbage transfer station to unload and then returns to the yard, and at the moment, the departure point in the travel and the yard of the garbage transfer station need to be removed;
and step 35, outputting the new individuals subjected to the enhanced local search strategy.
4. The garbage vehicle low-carbon scheduling method of claim 2, wherein the step of generating child particles by adaptively selecting a learning mode most conforming to the self stage of the particles by using a contribution-based adaptive learning strategy for the new individual comprises the following steps:
step 41, setting m learning strategies;
step 42, making contribution C of each learning strategykk＝0，kk＝1,2,…,m；
Step 43, selecting and determining a learning strategy, namely, learn (ii), for each new individual ii in the population through roulette respectively, generating new particles, namely, Npop (ii), according to the learn (ii), and calculating a target value of the Npop (ii);
step 44, sorting new particles in the new population according to the target value to obtain a ranking vector r (ii) of the particles;
step 45, assign a weight w to the ii new particle according to the following equationiiThe more the particle rank is, the larger the weight is distributed;
Figure FDA0003404383280000041
wherein wiiR (ii) a rank of the ii particle, as a weight to which the ii particle should be assigned;
and step 46, updating the contribution degree of the kth learning strategy according to the following formula,
Figure FDA0003404383280000042
wherein sfiiFor the reward factor, if the ii-th particle is promoted by the kth learning strategy fitness valuesfii1, otherwise sfii＝0；
Step 47, the method for updating and normalizing the kth learning strategy selection probability is shown as the following formula
pkk＝Ckk+ε kk＝1,2,...,m
Figure FDA0003404383280000051
Wherein p iskkSelecting probabilities for a kth learning strategy, jj representing subscripts of the learning strategies, wherein jj is 1, 2.
And 48, after the particles in the population are learned by selecting different learning modes, outputting all newly generated child particles Npop, the target value Npopbj of the new particle, the contribution degree vector C of the learning strategy and the selection probability vector P.
5. The garbage vehicle low-carbon scheduling method of claim 4, wherein the m learning strategies comprise: greedy learning strategies, multivariate learning strategies, exploratory learning strategies and exploratory learning strategies,
the greedy learning strategy is: generating a variation individual by reversing variation of the new individual, generating a cross individual I by the variation individual and an individual extreme value through a greedy cross operator, and performing greedy cross on the cross individual I and a global extreme value to generate a new individual after greedy learning;
the multivariate learning strategy is as follows: generating variant individuals by a new individual through a multivariate variant operator, carrying out partial mapping and crossing on the generated variant individuals and individual extreme values to obtain a cross individual I, and carrying out partial mapping and crossing on the cross individual 1 and a global extreme value to generate a new individual after multivariate learning;
the exploration type learning strategy is that new individuals generate variant individuals through a multivariate variant operator, and partial mapping and crossing are carried out on the generated variant individuals and the individual extreme value of any particle with fitness superior to the population, so as to generate new individuals after exploration and learning;
the utilization type learning strategy is that new individuals generate variant individuals through a multivariate mutation operator, and the generated variant individuals and individual extreme values are subjected to partial mapping and crossing to generate new individuals after utilization learning.
6. The garbage vehicle low-carbon dispatching method of claim 5, wherein the greedy learning strategy comprises:
step 411, determining the individuals X and the individual extremum X which need to be crossedpbestAnd global extremum Xgbest；
Step 412, randomly selecting one point from the points which are not the parking lot and the non-transfer station as a first garbage throwing point S of the vehicle service, and selecting the first garbage throwing point S at XpbestLeft garbage throwing point SLpAnd a right garbage throwing point SRpIn X, the left garbage throwing point SLXAnd a right garbage throwing point SRXAs a candidate set for the next access;
step 413, in candidate set { SLp,SRp,SLX,SRXSelecting a garbage throwing point S 'from the previous step, so that the S' and the S form a path section with the least transportation cost;
step 414, if S' ∈ { S ∈ [ ]Lp,SLXExecuting step 415, otherwise executing step 416;
step 415, from the individual extremum XpbestAnd deleting S from the individual X, setting S' as the next starting point S, and setting S as the left candidate set { SLp,SLXSelecting new S 'as a next garbage putting point for service, enabling S' and S to form a path section with the least transportation cost, and repeatedly executing S515 until a new solution X is generatednew；
Step 416, from the individual extremum XpbestAnd deleting S from the individual X, setting S' as the next starting point S, and setting S as the right candidate set { SRp,SRXSelecting a new S 'as a next garbage throwing point for service, so that the S' and the S form a path section with the least transportation cost, and repeatingS516 is executed until a new solution X is generatednew；
Step 417, with X and XpbestTake greedy intersection of (a) as an example, solve the new solution XnewAnd global extreme value XgbestPerforming greedy intersection again to generate a new solution;
step 418, output new individuals through greedy crossover operators.
7. The garbage vehicle low-carbon scheduling method of claim 5, wherein the multivariate learning strategy comprises:
step 421, calculating the selection probability of the reverse variation, the crossover variation and the insertion variation;
step 422, selecting the serial number of the variation mode based on the roulette mode;
step 423, if the sequence number of the reverse variation is selected, the individual is subjected to exchange variation;
step 424, if the sequence number of the crossover variation is selected, the individual is subjected to reverse variation;
step 425, if inserting the mutated serial number, performing insertion mutation on the individual;
and 426, outputting the new individuals passing through the multivariate mutation operator according to the mutation result.
8. The utility model provides a garbage vehicle low carbon dispatch system which characterized in that includes:
the acquisition module is used for acquiring input information and comprises: the number N of garbage throwing points needing to be served by the garbage vehicle, the coordinate information of the garbage throwing points, the coordinate information of a parking lot, the coordinate information of a garbage transfer station, the garbage amount of each garbage throwing point, the capacity Q of the garbage vehicle and the maximum working time T of a drivermax；
The determining module is used for inputting input information into a pre-constructed garbage vehicle low-carbon scheduling optimization model based on the adaptive particle swarm optimization algorithm and determining an optimal scheduling scheme;
the optimization target of the garbage vehicle low-carbon scheduling optimization model based on the adaptive particle swarm optimization is that the total transportation cost of vehicles in a planned scheduling scheme is minimum, the constraint conditions of the garbage vehicle low-carbon scheduling optimization model based on the adaptive particle swarm optimization are that all vehicles start from a parking lot and only start once, each garbage throwing point only allows one vehicle to serve once, when each garbage throwing point is served, one vehicle must drive from a certain place to the garbage throwing point and leave from the garbage throwing point, all vehicles empty garbage completely at a garbage transfer station, the garbage loading capacity of the vehicles in one stroke is not greater than the capacity limit of the vehicles, and the working time of a driver of each vehicle cannot exceed the specified maximum working time limit.
9. A computer readable storage medium storing one or more programs, wherein the one or more programs comprise instructions, which when executed by a computing device, cause the computing device to perform any of the methods of claims 1-7.
10. A computing device, comprising:
one or more processors, memory, and one or more programs stored in the memory and configured to be executed by the one or more processors, the one or more programs including instructions for performing any of the methods of claims 1-7.
CN202111508631.0A 2021-12-10 2021-12-10 Garbage vehicle low-carbon scheduling method and system Pending CN114611864A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111508631.0A CN114611864A (en) 2021-12-10 2021-12-10 Garbage vehicle low-carbon scheduling method and system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111508631.0A CN114611864A (en) 2021-12-10 2021-12-10 Garbage vehicle low-carbon scheduling method and system

Publications (1)

Publication Number Publication Date
CN114611864A true CN114611864A (en) 2022-06-10

Family

ID=81857916

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111508631.0A Pending CN114611864A (en) 2021-12-10 2021-12-10 Garbage vehicle low-carbon scheduling method and system

Country Status (1)

Country Link
CN (1) CN114611864A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117094534A (en) * 2023-10-19 2023-11-21 浩博泰德(北京)科技有限公司 Intelligent control method and system for Internet of things

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117094534A (en) * 2023-10-19 2023-11-21 浩博泰德(北京)科技有限公司 Intelligent control method and system for Internet of things
CN117094534B (en) * 2023-10-19 2024-01-23 浩博泰德(北京)科技有限公司 Intelligent control method and system for Internet of things

Similar Documents

Publication Publication Date Title
CN107092976A (en) A kind of method of multi-objective Model collaboration optimization a plurality of public bus network departure interval
CN110189025B (en) Electric vehicle charging station planning scheme acquisition method considering different load increases
CN108805335B (en) Public bicycle scheduling method
CN110516870B (en) Multi-recovery site garbage collection and transportation method based on co-evolution
CN112434854B (en) Intelligent environmental sanitation management method and device, computer equipment and readable medium
CN116127857B (en) Classification-oriented household garbage collection and transportation path multi-objective optimization method and system
CN113326993A (en) Shared bicycle scheduling method based on deep reinforcement learning
CN112097783B (en) Electric taxi charging navigation path planning method based on deep reinforcement learning
CN115146946A (en) Automobile charging facility selection and planning method based on multi-source data
CN113256069A (en) Hybrid bus fleet scheduling method based on power energy difference
CN114611864A (en) Garbage vehicle low-carbon scheduling method and system
CN112116125A (en) Electric vehicle charging navigation method based on deep reinforcement learning
Shen et al. Energy-efficient multi-trip routing for municipal solid waste collection by contribution-based adaptive particle swarm optimization
CN110674988A (en) Urban charging station planning method based on electric vehicle travel big data
Ye et al. Multi-objective optimisation for taxi ridesharing route based on non-dominated sorting genetic algorithm
Moyo et al. Advancements in E-mobility: A bibliometric literature review on battery technology, charging infrastructure, and energy management
CN116415745A (en) Low-carbon dynamic path planning method and system for garbage vehicle
CN112561104A (en) Vehicle sharing service order dispatching method and system based on reinforcement learning
CN112434881B (en) Charging station position screening method and system
CN115099702B (en) Electric bus daytime running charging optimization method based on Lagrange relaxation algorithm
CN116804559A (en) Medical waste clearing method based on interval multi-target mixed frog-leaping algorithm
CN115063013A (en) Receiving and transporting scheduling method, system and medium based on renewable resources
CN117830062B (en) Household garbage collection and transportation system planning method based on full life cycle carbon emission accounting
Guang et al. Location planning of charging station for electric vehicle based on urban traffic flow
Zhu Mobile battery replacement service routing problem for shared electric bikes based on electric logistics vehicles

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination