Oil port resource optimization scheduling method

Abstract

The present invention is an oil port resource optimization scheduling method, which includes the following steps: step S11 of establishing a lower-level scheduling model; step S12 of establishing an upper-level scheduling model; step S13 of obtaining the lower-level scheduling model and the upper-level scheduling model using a multi-objective evolutionary algorithm; and step S14 of determining the target berth and berthing sequence of the tanker based on the obtained results. According to the present invention, on the one hand, under the factor conditions of multiple storage tanks and multiple pipelines, the operation of the entire oil port can be relatively balanced and relatively stable, and on the other hand, the control objective of maximizing the profit of the oil port terminal and minimizing the tanker's stay time in port can be achieved. The present invention is an optimal scheduling solution, which is obviously improved in both intelligence and streamlining compared with the traditional manual creation method.

Description

[Technical field]

[0001] The present invention belongs to the technical field of port facilities, and particularly relates to an oil port resource optimization scheduling method. [Background technology]

[0002] An oil port is a dedicated port for loading and unloading crude oil and refined petroleum products. Oil port resource scheduling is of great significance to the construction of smart ports. Reasonable scheduling can equalize berth occupancy time, improve the profits of oil ports, and optimize the work efficiency of oil ports.

[0003] At present, oil port resource scheduling generally adopts the form of manual unloading planning: before the tanker enters the port, the land side provides the ship side with information such as the number of berths and pipelines connected between the ship and the land side, the ship side provides the ship side with information such as the oil type and tonnage to be loaded during the voyage, and both the ship side and the land side confirm the information such as the type and amount of crude oil, the unloading order and the unloading rate, and then formulate and execute the unloading plan. However, since multiple pipelines may be transported at the same time during the unloading operation and multiple tanks may be stored at the same time, the traditional manual unloading planning technology has difficulty in optimizing the oil port resource allocation, and the oil port resources cannot be fully utilized, which has problems to be improved, such as intelligence and rationalization. Summary of the Invention [Problem to be solved by the invention]

[0004] The present invention designs and provides a method for optimally scheduling oil port resources, in contrast to the existing manual unloading work planning technology of oil ports, which has difficulties in optimizing oil port resource allocation, cannot fully utilize oil port resources, and has problems in need of improvement in intelligence and rationalization. [Means for solving the problem]

[0005] To achieve the above object of the invention, the present invention is realized by using the following technical solutions.

[0006] In the oil port resource optimization scheduling method, It comprises the steps of: A step S11 of establishing a lower-level scheduling model, The lower level scheduling model is as follows:

number

number

[0007] Compared with the prior art, the advantages and positive effects of the present invention are as follows: on the one hand, the present invention can relatively balance and stabilize the work of the entire oil port under the factor conditions of multiple storage tanks and multiple pipelines, and on the other hand, it can achieve the control objective of maximizing the profit of the oil port terminal and minimizing the tanker's stay time in port. The present invention is an optimal scheduling solution, and compared with the traditional manual preparation method, both intelligence and streamlining are obviously improved. [Brief description of the drawings]

[0008] In order to more clearly describe the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings necessary for the embodiments. Obviously, the drawings in the following description are some embodiments of the present invention, and those skilled in the art can obtain other drawings from these drawings without creative efforts.

[0009] [Figure 1]FIG. 2 is a schematic diagram of an oil port structure applied to the resource optimization scheduling method proposed in the present invention; [Diagram 2] 1 is a flowchart of an oil port resource optimization scheduling method proposed by the present invention; [Diagram 3] 1 is a flowchart for solving a lower-level scheduling model and an upper-level scheduling model using a multi-objective evolutionary algorithm. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0010] In order to make the objectives, technical solutions and advantages of the present invention clearer and easier to understand, the present invention is further described in detail in the following combined with the accompanying drawings and embodiments.

[0011] In describing the present invention, terms indicating directions or positions, such as "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc., are based on the directions or positions shown in the accompanying drawings, and are merely for convenience of description, and do not indicate or imply that the devices or elements must be configured or operated in a particular direction, and therefore are not to be construed as limitations of the present invention. Furthermore, the terms "first" and "second" are used for descriptive purposes only, and are not to be understood as indicating or implying relative importance.

[0012] FIG. 1 is a schematic diagram showing the structure of an oil port applied to the oil port resource optimization scheduling method proposed by the present invention. As shown in FIG. 1, when performing unloading work, the oil port resource can be divided into a tanker layer, a berth layer, and a tank layer. In FIG. 1, the tanker layer includes, for example, tanker 1, tanker 2, ..., tanker n, the berth includes, for example, berth 1, berth 2, ..., berth m, and the tank includes, for example, tank 1, tank 2, ..., tank r. The berth layer and the tank layer are communicated through multiple pipelines, and when unloading the same type of oil, multiple pipelines can be transported in parallel at the same time. The unloaded oil can be stored in target tanks installed according to the type of oil, and each tank is used to store one type of oil, and each tank is installed arbitrarily in multiple numbers according to demand.

[0013] The oil port resource optimization scheduling method proposed in the present invention is intended to allocate berths, and the incoming tanker selects berths according to the efficient principle, and allocates berths to the tanker in an optimal order so that the tanker can berth and perform unloading work, control the total residence time of the tanker in the oil port, and improve the profit of the oil port. Berth allocation is the basis of oil port resource scheduling and the core of the entire optimization scheduling method.

[0014] Berth allocation has the following characteristics: (1) After a tanker arrives at a port, a determination is made as to whether or not there is a vacant berth. If there is a vacant berth, the next berth is allocated; if there is no vacant berth, the tanker waits at the anchorage. (2) Only one tanker may dock at a berth at any one time and one tanker may occupy only one berth. (3) After the tanker has completed the unloading process, it is allowed to leave the port and one unloading operation is completed. Once the tanker has left the port and returned to the port to berth, it is certified to be ready for the next unloading operation. (4) When multiple types of oil are stored in one tanker, they can only be unloaded sequentially; multiple types of oil cannot be unloaded at the same time. (5) Factors that may cause delays in the arrival of a tanker due to force majeure, such as weather or equipment failure, will not be taken into consideration.

[0015] From the schematic diagram of the oil port structure in Figure 1 and the characteristics of berth allocation, when unloading crude oil, multiple pipelines can transport oil at the same time, multiple tanks can store oil at the same time, and different tanks can store different types of oil. Therefore, the amount of oil transported by the pipeline affects the oil transport time, and the larger the amount of oil transported by the pipeline, the shorter the oil transport time. Conversely, the smaller the amount of oil transported by the pipeline, the longer the oil transport time. On the other hand, the berth and berthing sequence at which the tanker is anchored affect the berth occupation period and oil transport time, and if the berth and berthing sequence at which the tanker is anchored are reasonable, the berth occupation period and oil transport time will be shortened. Conversely, if the berth and berthing sequence at which the tanker is anchored are not reasonable, the berth occupation period and oil transport time will be lengthened. Generally, the total dwell time of a tanker is the sum of the tanker's waiting time at berth, the time to prepare for crude oil transportation, and the time to transport the crude oil. A rational berth allocation can minimize the tanker's waiting time, the time to prepare for crude oil transportation, and the time to transport the crude oil, shorten the total dwell time of the tanker, improve the effective unloading operation time of the berth, and improve the profits of the oil port.

[0016] In order to achieve the above object, the oil port resource optimization scheduling method provided by the present invention includes a number of steps, as detailed below: The oil port resource optimization scheduling method provided by the present invention may be executed in an oil port resource management system, and the oil port resource management system may be realized by a controller having a processor, and the controller further includes a storage device and necessary peripheral circuits connected to the controller.

[0017] Step S11: Establishing a lower-level scheduling model is step S11. The control objectives of the lower-level scheduling model include the maximum berth utilization equalization rate (minimum standard deviation) and the maximum pipeline transportation time equalization rate. The low-level scheduling model can be expressed as follows:

number

[0018] The lower-level scheduling model obtains a feasible solution, i.e., an ideal tanker berth, in which the berth occupancy time and the transportation time through the pipeline are kept in equilibrium, i.e., in a multi-factor condition of multiple tanks and multiple pipelines, the tanker, berth, pipelines and tanks perform unloading operations in an interrelated state, and are in a relative equilibrium and relative stable state.

[0019] Considering the characteristics of unloading operations at oil ports, the lower-level scheduling model further includes the following constraints: (1) The tanker berthing constraints are that a tanker can only perform unloading operations when it enters an oil port, and that a tanker can only be berthed at one berth. The tanker berthing constraints can be expressed by the following equation. That is, JPEG2025502595000024.jpg10136Here, i is a tanker, i = 1, 2, · · · , n, and n denotes the number of tankers. (2) The tanker load constraint is that the tanker is allowed to enter the target berth only if the tanker's load is smaller than the berthing capacity of the target berth. The tanker load constraint is expressed by the following formula, i.e., JPEG2025502595000025.jpg5134Where, i denotes the loading capacity of tanker i, and W k indicates the berthing capacity of the berth. In the tanker berthing constraint and tanker load weight constraint, y i,k satisfies the following formula: JPEG2025502595000026.jpg14137(3) The pipeline constraint is that at least one pipeline is installed between the target berth and the tank to realize the oil unloading operation. The pipeline constraint can be expressed as follows: JPEG2025502595000027.jpg11135Here, q is the number of pipelines, j is the oil type, j = 1, 2, · · ·, u, u is the number of oil types, s is the tank, s = 1, 2, · · ·, r, r is the number of tanks. JPEG2025502595000028.jpg12138(4) The unloading capacity constraint condition is that during the set planning period, the oil type to be unloaded must be consistent with the oil type that can be accommodated in the target tank, and the target unloading amount of oil in the tanker is less than the remaining capacity of the target tank. The unloading capacity constraint can be expressed as follows, i.e., JPEG2025502595000029.jpg12132, where F s denotes the remaining storage capacity of tank s. JPEG2025502595000030.jpg13137, where E i,j denotes the tonnage of the jth oil unloaded by tanker i. The tonnage of the jth oil unloaded by tanker i is a known value and can be obtained by scheduling from the tanker side. The remaining storage capacity of tank s can be detected in real time, for example, by detecting the real-time liquid level of tank s using a radar level sensor. (5) The unloading target constraint is that the oil in the tanker needs to be completely unloaded. The unloading target constraint can be expressed as follows: JPEG2025502595000031.jpg10136, where E i,j,p denotes the tonnage of the jth oil to be transported to tanker i via pipeline p.

[0020] The feasible solution obtained by the lower-level scheduling model, i.e., the berth where the tanker is docked, the determined oil storage tank, the pipeline, and the pipeline transportation volume, can ensure the effective implementation of the unloading operation, and at the same time, can ensure that the berth occupation time and the pipeline transportation time are kept in equilibrium, and can ensure that the entire operation of the oil port is effective and stable, and in relative equilibrium and relative stability. Based on the lower-level scheduling model, a higher-level scheduling model is further constructed.

[0021] Step S12: The oil port resource optimization scheduling method provided by the present invention includes the following steps: Step S12: Establish an upper-level scheduling model. The control objective of the upper-level scheduling model is to maximize the profitability of the oil port terminal and minimize the total residence time of the tanker. The high level scheduling model is given by the following equation:

number

[0022] The feasible solution obtained from the high-level scheduling model, i.e., the tanker berthing sequence, can maximize the oil port terminal's profit and minimize the tanker's total residence time.

[0023] Considering the characteristics of unloading work at the oil port, the upper-level scheduling model further includes the following constraints: (1) The pipeline constraint is that at least one pipeline is installed between the target berth and the tank to realize the oil unloading operation. The pipeline constraint can be expressed as follows: JPEG2025502595000034.jpg12136 where k is the number of bars, k = 1, 2, . . . , m, where m is the number of berths of the oil port. p is the pipeline, p=1,2,···,q, where q is the number of pipelines. j is the oil type, j = 1, 2, , u, where u is the number of oil types. JPEG2025502595000035.jpg14144(2) The berth access permission constraint is that only one tanker can be anchored at one berth. The berth access permission constraint can be expressed as follows: JPEG2025502595000036.jpg6136, where TB i+1.kindicates the time when the i+1th tanker arrives at berth k, and TL i,k indicates the time when tanker i completes work at berth k and departs, i.e., the time when the (i+1)th tanker docks at berth k is later than the time when the ith tanker completes work at berth k and departs. (3) The unloading completion departure constraint is that the tanker is allowed to leave the port after completing the unloading operation, and one unloading operation is completed. The unloading completion departure constraint can be expressed as follows. That is, JPEG2025502595000037.jpg12134 Here, TL i,k indicates the time when tanker i completes its work at berth k and leaves the port, TB i,k indicates the time when tanker i will berth k, JPEG2025502595000038.jpg7136 shows the maximum value of the sum of the preparation time and transportation time when transporting the jth oil of tanker i through a pipeline. The preparation time TR for the jth oil shipment of tanker i through pipeline p i,j,p is calculated by the following formula: JPEG2025502595000039.jpg6136, where TR p indicates the preparation time for transporting oil through the pipeline p, which includes, but is not limited to, the start-up preparation time of fluid control components such as pumps and valves installed in the pipeline p, and TR p may be generated based on detected historical data and pre-stored in a storage device for retrieval on demand. The transit time for transporting the jth oil of tanker i through pipeline p is calculated by the following formula: JPEG2025502595000040.jpg11143, where E i,j,p denotes the tonnage of oil j transported by tanker i through pipeline p, and v j,p denotes the flow rate of the jth oil transported through pipeline p. In the above two formulas, W i,j,p satisfies the following formula: That is, when the jth oil of the tanker i is transported to the tank through the pipeline p, W i,j,p is 1. Otherwise, W i,j,p is 0.

[0024] Step S13: Step S13 is a step of obtaining a lower-level scheduling model and an upper-level scheduling model by using a multi-objective evolutionary algorithm. The lower-level scheduling model and the upper-level scheduling model are obtained by software, such as Matlab. In step S14, a target berth and a berthing sequence for the tanker are determined based on the obtained results.

[0025] In an optional embodiment, a multi-objective evolutionary algorithm is used to solve the lower-level scheduling model. Optionally, solving the lower-level scheduling model using a multi-objective evolutionary algorithm includes the following steps. That is, Step S21: An individual encoding, in binary form, of a birth number corresponding to a feasible solution of the lower-level scheduling model. In this embodiment, a berth where a tanker is moored, which corresponds to a feasible solution (also called an individual) of the lower-level scheduling model, has a unique berth number, which is represented by a binary number. The berth number represented by a binary number constitutes the phenotype of the individual (i.e., the phenotype of the feasible solution). The phenotype and the berth number can be mutually converted by an encoding and decoding program. Step S22: Generate an initial population. The size of the initial population is set, and individuals of the initial population are randomly selected in the feasible solution of the lower-level scheduling model, and the initial population is defined as pop1. Step S23: A fitness calculation step S23 for calculating the fitness of the sample individuals. Since each objective in the low-level scheduling model does not satisfy linear independence, it is preferable to construct the fitness function using the Technique for Order Preference by Similarity to an Ideal Solution (TOPSIS). The fitness function is shown as follows: JPEG2025502595000042.jpg11132, where g l denotes the fitness value of the l-th individual in the initial population of the lower-level scheduling model, and g l The smaller the value, the better the individual. JPEG2025502595000043.jpg6137 shows the distance from the objective function value of the first individual in the initial population of the lower-level scheduling model to the negative ideal solution of the higher-level scheduling model. JPEG2025502595000044.jpg6137 shows a distance from the objective function value of the first individual of the initial population of the lower-level scheduling model to the positive ideal solution of the lower-level scheduling model. The objective function is selectably set to the model function of the lower-level scheduling model. Step S24: A selection operation is performed based on the fitness of the sample individuals to select the individual sample in a selection operation step S24. Calculate the sum of the fitnesses of all individuals in the initial population. Calculate the relative fitness of each individual in the initial population, where the relative fitness is the ratio of the individual fitness to the sum of the fitnesses. Each relative fitness can be represented by a probability, and since the sum of all probability expressions is 1, it can be considered that the region to which the probability values ​​corresponding to all relative fitnesses are assigned can be formed into a complete region (e.g., a disk). Selection markers are randomly placed in this complete region, and the selected independent regions are the selected individuals. That is, individuals are selected from the population in a manner similar to spinning a disk. A target number of individuals are selected from the initial population by spinning a target sub-disk, i.e., in the standard Roulette wheel manner. If the fitnesses of the individuals in a step are normalized, they can also be used directly as the roulette wheel probabilities. Step S25: An intersection operation step S25 in which the individual samples after the selection operation are subjected to an intersection operation. Randomly select two individuals from the selected individual sample and perform crossover with a given crossover probability. Linearly interpolate the factors corresponding to the random positions with the coefficients, and leave the remaining positions unchanged to obtain two new individuals. The crossover probability is adjusted according to the fitness of the two individuals involved in the crossover operation. First, the crossing probability interval [pc min ,pc max ] and the individual fitnesses of the population, the average fitness f avg , the minimum fitness f min The crossover probability is determined by the following formula: JPEG2025502595000045.jpg23136Here, f' is the greater fitness of the two individuals involved in the crossover operation. Step S26: A mutation operation step S26 in which the individual samples after the crossover operation are mutated. According to the variation probability, the values ​​of one or more variation points in the post-crossover individual sample are inverted. First, the interval of mutation probability [pm min ,pm max] is set and the mutation probability is adjusted according to the individual fitness. The mutation probability is determined by the following formula: JPEG2025502595000046.jpg23135Here, pm denotes the mutation probability, pm min indicates the minimum value of the set mutation probability interval, and pm max indicates the maximum value of the set mutation probability interval, and f min indicates the fitness minimum in the population of sample individuals after crossover, and f avg denotes the average fitness in the population of sample individuals after crossover, and f denotes the fitness value of the mutant individuals. Step S27: In step S27, judge whether the iteration condition is satisfied. If the iteration condition is satisfied, terminate the calculation and obtain the optimized population pop1-opt of the lower-level scheduling model. If the iteration condition is not satisfied, take the individual sample after the mutation operation as the population of the next generation, and repeat steps S23 to S26 until the iteration condition is satisfied. Specifically, the next generation population is obtained through selection, crossover, and mutation operations. Furthermore, the selection, crossover, and mutation operations are repeated until the number of iterations reaches the upper limit (the iteration condition is satisfied), and the calculation is terminated to obtain the optimized population pop1-opt corresponding to the maximum fitness of the lower-level scheduling model. The factor type of the individuals in the optimized population pop1-opt is decoded, that is, the berth where the oil port is moored is obtained, and the oil storage tank, the oil transport pipeline, and the pipeline transport volume are determined. Step S28: Step S28 is a step of dividing E corresponding to each individual in the optimization population pop1-opt. i,j,p Calculate. E i,j,p denotes the tonnage of oil j from tanker i transported through pipeline p. Step S29: Step S29 is a step of i,j,p into the high-level scheduling model. Multi-objective evolutionary algorithms are i,j,p is used to solve the higher-level scheduling model into which

[0026] If necessary, a multi-objective evolutionary algorithm is used to i,j,p Solving the high-level scheduling model with That is, Step S30: An individual encoding step S30 indicates, in binary form, the birth number corresponding to the feasible solution of the upper level scheduling model. In this embodiment, a tanker berth corresponding to a feasible solution (also called an individual) of the upper-level scheduling model has a unique berth number, which is represented by a binary number. The binary berth number constitutes the phenotype of the individual (i.e., the phenotype of the feasible solution). The phenotype and the berth number can be mutually converted by an encoding and decoding program. Step S31: Generate an initial population. The size of the initial population is set, and individuals of the initial population are randomly selected in the feasible solution of the upper-level scheduling model, and the initial population is defined as pop2. Step S32. A fitness calculation step S32 for calculating the fitness of the sample individuals. Since each objective in the high-level scheduling model does not satisfy linear independence, it is preferable to construct the fitness function using the Technique for Order Preference by Similarity to an Ideal Solution (TOPSIS). The fitness function is shown as follows: JPEG2025502595000047.jpg11135, where g l denotes the fitness value of the I-th individual in the initial population of the upper-level scheduling model, and g l The smaller the value, the better the individual. JPEG2025502595000048.jpg6137 shows the distance from the objective function value of the first individual in the initial population of the higher-level scheduling model to the negative ideal solution of the higher-level scheduling model. JPEG2025502595000049.jpg6136 shows the distance from the objective function value of the first individual of the initial population of the higher-level scheduling model to the positive ideal solution of the higher-level scheduling model. The objective function is selectable and set to the model function of the higher-level scheduling model. For a function that takes a maximum value, it can be converted to take a minimum value by inverting the numerical form, and the overflow of the function can be prevented by adding 1 to the denominator. Step S33: A selection operation is performed based on the fitness of the sample individuals to select the individual sample in a selection operation step S33. Calculate the sum of the fitnesses of all individuals in the initial population. Calculate the relative fitness of each individual in the initial population, where the relative fitness is the ratio of the individual fitness to the sum of the fitnesses. Each relative fitness can be represented by a probability, and since the sum of all probability expressions is 1, it can be considered that the region to which the probability values ​​corresponding to all relative fitnesses are assigned can be formed into a complete region (e.g., a disk). Selection markers are randomly placed in this complete region, and the selected independent regions are the selected individuals. That is, individuals are selected from the population in a manner similar to spinning a disk. A target number of individuals are selected from the initial population by spinning a target sub-disk, i.e., in the standard Roulette wheel manner. If the fitnesses of the individuals in a step are normalized, they can also be used directly as the roulette wheel probabilities. Step S34: An intersection operation step S34 in which the individual samples after the selection operation are subjected to an intersection operation. Randomly select two individuals from the selected individual sample and perform crossover with a given crossover probability. Linearly interpolate the factors corresponding to the random positions with the coefficients, and leave the remaining positions unchanged to obtain two new individuals. The crossover probability is adjusted according to the fitness of the two individuals involved in the crossover operation. First, the crossing probability interval [pc min ,pc max ] and the individual fitnesses of the population, the average fitness f avg , the minimum fitness f min The crossover probability is determined by the following formula: JPEG2025502595000050.jpg23136Here, f' is the greater fitness of the two individuals involved in the crossover operation. Step S35: A mutation operation step S35 in which the sample individuals after the crossover operation are mutated. According to the variation probability, the values ​​of one or more variation points in the post-crossover individual sample are inverted. First, the interval of mutation probability [pm min ,pm max ] is set, and the mutation probability is adjusted according to the individual fitness. The mutation probability is determined by the following formula: JPEG2025502595000051.jpg23136Here, pm denotes the mutation probability, pm min indicates the minimum value of the set mutation probability interval, and pm max indicates the maximum value of the set mutation probability interval, and f min indicates the fitness minimum in the population of sample individuals after crossover, and f avg denotes the average fitness in the population of sample individuals after crossover, and f denotes the fitness value of the mutant individuals. Step S36: In step S36, judge whether the iteration condition is satisfied. If the iteration condition is satisfied, terminate the calculation and obtain the optimized population pop1-opt of the upper level scheduling model. If the iteration condition is not satisfied, take the individual sample after the mutation operation as the population of the next generation, and repeat steps S32 to S35 until the iteration condition is satisfied.

[0027] The factor types of the individuals in the optimization population pop2-opt are decoded to obtain the berthing sequence of the oil ports. By solving the genetic algorithm twice and obtaining the optimal solution of the upper-level scheduling model based on the optimal solution of the lower-level scheduling model, the berths, oil storage tanks, oil pipelines, pipeline transport volume, and berthing sequence of the oil port can finally be obtained. Under the conditions of multiple storage tanks and multiple pipelines, it can ensure that the operation of the entire oil port is relatively balanced and stable, while achieving the control goal of maximizing the profit of the oil port terminal and minimizing the tanker's stay time in port, which is an optimized scheduling plan, which is obviously improved in intelligence and rationality compared with the traditional manual preparation method.

[0028] The above embodiments are only used to explain the technical solutions of the present invention, but are not used to limit them. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art can still make modifications to the technical solutions described in the above embodiments or make equivalent replacements for some technical features therein, and such modifications or replacements do not deviate from the essence of the corresponding technical solutions from the spirit and scope of the technical solutions protected by the present invention.

Claims

1. In the oil port resource optimization scheduling method, It comprises the steps of: A step S11 of establishing a lower level scheduling model, The low-level scheduling model is as follows: [0010] Where: k is a berth, k=1, 2, ..., m, m is the number of berths of the oil port, T ｋ is the usage time of berth k within the set planning period, [0025] Where: 【number】 is the average usage time of all m berths within the set planning period, [0030] TR ｉ，ｊ，ｐ denotes the preparation time for transporting the jth oil of tanker i through pipeline p, p is a pipeline, p=1, 2, ..., q, q is the number of pipelines, j is an oil type, j=1, 2, ..., u, u is the number of oil types, i is a tanker, i=1, 2, ..., n, n is the number of tankers; TR ｉ，ｊ，ｐ =w ｉ，ｊ，ｐ TR ｐ , T.R. Ｐ is the preparation time for transporting oil through pipeline p, and w ｉ，ｊ，ｐ satisfies the following formula, that is, 【number】 TZ ｉ，ｊ，ｐ denotes the transportation time of oil j of tanker i transported through pipeline p, [0045] where E ｉ，ｊ，ｐ denotes the tonnage of oil j of tanker i transported through pipeline p, and v ｊ、ｐ denotes the flow rate of the jth oil transported through the pipeline p, 【number】 When the condition for transporting the jth oil of tanker i through pipeline p is satisfied, the sum of the preparation time and the transportation time is TR ｉ，ｊ，ｐ +TZ ｉ，ｊ，ｐ Show that 【number】 denotes the maximum sum of preparation time and transportation time when transporting the jth oil of the tanker i through the pipeline p, 【number】 Step S11, where p indicates the minimum value of the sum of the preparation time and the transportation time when transporting the j-th oil of the tanker i through the pipeline p; A step S12 of establishing a high-level scheduling model, The high-level scheduling model is as follows: [0050] Here, u ｉ，ｊ denotes the price per ton paid by tanker i to unload the jth oil, E ｉ，ｊ denotes the tonnage of the jth oil unloaded by tanker i, 【number】 and L Ｐ denotes the length of the pipeline p, PC ｊ，ｐ denotes the unit cost of transporting the jth oil through pipeline p, E ｉ，ｊ，ｐ denotes the tonnage of the jth oil of tanker i transported through pipeline p, DC ｉ indicates the penalty cost for the tanker i's dwell time, TL ｉ，ｋ indicates the departure time of tanker i after the work at berth k is completed, T.D. ｉ denotes the latest allowable departure time for tanker i, T.A. ｉ indicates the arrival time of the tanker i, step S12; A step S13 of determining the lower-level scheduling model and the upper-level scheduling model using a multi-objective evolutionary algorithm; Step S14: determining a target berth and a berthing sequence for the tanker based on the obtained results; An oil port resource optimization scheduling method comprising:

2. The lower-level scheduling model further includes the following constraints: The constraints for a tanker to berth are as follows: [006] and The tanker load constraints are: [0070] And W ｉ indicates the loaded weight of tanker i, and W ｋ indicates the berthing capacity of berth k, The pipeline constraints are: [0080] where s is a tank, s=1, 2, ..., r, r is the number of tanks, 【number】 and The constraints on the unloading capacity are: [0097] where F ｓ is the remaining capacity of tank s, 【number】 and E ｉ，ｊ is the tonnage of oil j unloaded from tanker i, The unloading goal constraints are: [0089] The oil port resource optimization scheduling method according to claim 1,

3. The lower-level scheduling model further includes the following constraints: The pipeline constraints are: ##EQU00011## and Where: 【number】 and The permission constraints for berth approval are: ##EQU00012## where: T.B. ｉ＋１，ｋ indicates the time when the i+1th tanker arrives at berth k, TL ｉ，ｋ indicates the departure time of tanker i after the work at berth k is completed, The departure constraints for unloading completion are: ##EQU00013## and Here, T.L. ｉ，ｋ indicates the departure time of tanker i after the work is completed at berth k, and TB ｉ，ｋ indicates the time when the i-th tanker arrives at berth k, 【number】 denotes the maximum value of the sum of the preparation time and the transportation time for transporting the jth oil of the tanker i through the pipeline p. The oil port resource optimization scheduling method according to claim 1 or 2.

4. Obtaining the lower-level scheduling model and the upper-level scheduling model using the multi-objective evolutionary algorithm includes the following steps: Step S21: indicating, in binary, a berth number corresponding to a feasible solution of the lower-level scheduling model; A step S22 of setting a size of an initial population, randomly selecting individuals of the initial population in a feasible solution of the lower-level scheduling model, and defining the initial population as pop1; A step S23 of calculating the fitness of the sample individuals; A step S24 of performing a selection operation based on the fitness of the sample individuals to select an individual sample; A step S25 of performing an intersection operation on the individual samples after the selection operation; A step S26 of performing a mutation operation on the individual sample after the crossover operation; Step S27: determining whether an iteration condition is satisfied, and if the iteration condition is satisfied, terminating the calculation and obtaining an optimized population pop1-opt of the lower-level scheduling model; if the iteration condition is not satisfied, using the individual sample after the mutation operation as the next generation population, and repeatedly executing steps S23 to S26 until the iteration condition is satisfied; E corresponding to each individual of the optimized population pop1-opt ｉ，ｊ，ｐ Calculating E ｉ，ｊ，ｐ step S28, where p denotes the tonnage of oil j of tanker i transported through pipeline p; The calculated E ｉ，ｊ，ｐ into the upper level scheduling model; a step S30 of representing in binary form tanker berthing sequences corresponding to feasible solutions of the upper level scheduling model; Step S31: setting an initial population size, randomly selecting individuals of the initial population in a feasible solution of the upper-level scheduling model, and defining the initial population as pop2; A step S32 of calculating the fitness of the sample individuals; A step S33 of performing a selection operation based on the fitness of the sample individuals to select individual samples; A step S34 of performing a crossover operation on the sample individuals after the selection operation; A step S35 of performing a mutation operation on the sample individual after the crossover operation; If the iteration condition is satisfied, the calculation is terminated, and the optimized population pop2-opt of the upper level scheduling model is obtained. If the iteration condition is not satisfied, the individual sample after the mutation operation is used as the next generation population, and steps S32 to S36 are repeatedly executed until the iteration condition is satisfied. and S37. decoding the optimized population of the upper level scheduling model to obtain tanker target berths and berthing sequences. The oil port resource optimization scheduling method according to claim 1 .

5. The fitness function when performing the fitness calculation in step S23 or step S32 is ##EQU00014## and Here, g ｌ denotes the fitness value of the first individual in the initial population of the upper-level scheduling model, and g ｌ The smaller the value, the better the individual. 【number】 indicates a distance from the objective function value of the first individual of the initial population of the higher-level scheduling model to the negative ideal solution of the higher-level scheduling model, 【number】 indicates the distance from the objective function value of the first individual of the initial population of the higher-level scheduling model to the positive ideal solution of the higher-level scheduling model, When performing the fitness calculation in step S23, the objective function is the lower-level scheduling model, and when performing the fitness calculation in step S32, the objective function is the upper-level scheduling model. The oil port resource optimization scheduling method according to claim 4 .

6. When the crossover calculation is performed in step S25 or step S34, the crossover probability is determined by the following formula: ##EQU00015## where p is the crossover probability, and p ｍｉｎ is the minimum value of the set crossing probability interval, and p ｍａｘ is the maximum value of the set crossover probability interval, f′ is the individual with the greater fitness among the two individuals involved in the crossover operation, and f ａｖｇ The oil port resource optimization scheduling method according to claim 4, characterized in that: is the average fitness.

7. When the mutation operation is performed in step S26 or step S35, the mutation probability is determined by the following formula: ##EQU00016## where pm is the mutation probability and pm ｍｉｎ indicates the minimum value of the set mutation probability interval, and pm ｍａｘ indicates the maximum value of the set mutation probability interval, and f ｍｉｎ indicates the fitness minimum in the population of sample individuals after the crossover operation, and f ａｖｇ The oil port resource optimization scheduling method according to claim 4, characterized in that f denotes the fitness average value in the population of the sample individuals after the crossover operation, and f denotes the fitness value of the mutant individuals.

Images