Coal port loading and unloading production intelligent scheduling implementation method, system, equipment and medium

Abstract

The invention provides an intelligent scheduling implementation method for coal port loading and unloading production. The method comprises the steps that a loading and unloading mixed scheduling optimization model and an improved genetic algorithm are set; solving the loading and unloading hybrid scheduling optimization model based on the improved genetic algorithm so as to obtain an optimal scheduling scheme; the improved genetic algorithm is improved through a multi-dimensional vector coding mechanism, population initialization based on a process constraint rule and a fitness function of a calling conflict punishment mechanism; the invention further provides a coal port loading and unloading production intelligent scheduling implementation system which comprises a loading and unloading mixed scheduling optimization model and an improved genetic algorithm. The invention further provides an electronic device and a computer readable storage medium applying the method. The method is suitable for scheduling optimization of southern water transfer type coal ports, the overall operation efficiency is improved, and the time of all the ships to be loaded and unloaded in the port is shortest in the operation plan period.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent scheduling of coal port loading and unloading production, and in particular to a method, system, equipment and medium for realizing intelligent scheduling of coal port loading and unloading production. Background Art

[0002] At present, intelligent scheduling systems in the port field are mainly used in automated container terminals, such as berth allocation of container terminals, path planning of AGV (automated guided vehicles) in the yard, and scheduling optimization of loading and unloading equipment (quay cranes). However, intelligent scheduling is less used in bulk terminals, especially coal terminals with specific cargo types. Manual scheduling is still widely used to arrange terminal production. Manual scheduling is difficult to make optimal scheduling decisions when faced with complex scenarios such as diversified ship types, nonlinear surge in port arrival frequency, and mixed loading of multiple coal types.

[0003] At the few bulk cargo terminals that utilize intelligent scheduling systems, the existing scheduling models do not consider the coordinated optimization of loading and unloading operations. Because the loading and unloading processes of ordinary bulk cargo ports are independent, these scheduling models only consider the scheduling optimization of a single link, such as berth scheduling, yard scheduling, or loading scheduling. Intelligent scheduling optimization of a single link can only achieve local optimization and cannot optimize overall port efficiency. Furthermore, the types of ship operations at the same type of ports vary across regions. For example, northern coal ports primarily export coal, with coal arriving by train and departing by ship. Therefore, northern coal ports primarily focus on loading operations. Southern coal ports, as coal transshipment hubs, primarily receive and unload large vessels from northern ports, then transport them to surrounding areas via inland waterways using smaller barges. Therefore, southern coal ports involve both loading and unloading operations. Clearly, single-link scheduling optimization cannot meet the scheduling optimization requirements of southern transshipment-type coal ports. For example, patent application publication number CN115456387A discloses a method, device, electronic device, and medium for integrated scheduling of coal port departure operations. The method includes: obtaining ship arrival information and port resource information; establishing, based on the ship arrival information and port resource information, a port integrated scheduling model with the goal of minimizing the total departure time of ships based on a basic MILP model; determining the constraints of the port integrated scheduling model; and using a preset optimization scheduling algorithm to calculate the optimal solution of the scheduling model under the constraints, and then obtaining an optimal port scheduling plan based on the optimal solution. The technical solution in this document is primarily applicable to northern coal ports that primarily handle ship loading.

[0004] Therefore, based on the needs of logistics efficiency and social benefits of southern coal ports, southern transit coal ports urgently need an efficient scheduling method to reasonably allocate berths, loading and unloading equipment, operating procedures, stacking spaces and other resources to each ship to maximize overall operating efficiency. Summary of the Invention

[0005] In view of this, it is necessary to propose methods, systems, equipment and media for intelligent scheduling of coal port loading and unloading production in response to the above-mentioned problems, so as to overcome some shortcomings in the above-mentioned background technologies and solve the following technical problems: How to realize intelligent scheduling of mixed loading and unloading production at the southern transit coal terminal, obtain the optimal scheduling plan, and thus maximize the overall operating efficiency of the coal terminal.

[0006] To achieve the above object, the present invention adopts the following technical solutions:

[0007] The present invention proposes a method for implementing intelligent scheduling of coal port loading and unloading production, which is applied to a coal port including at least two ship loaders, at least two ship unloaders, at least two stackers, at least two loading berths, and at least two unloading berths. The method comprises the following steps S1 to S2, which are performed in the following order:

[0008] Step S1, setting a mixed loading and unloading scheduling optimization model and an improved genetic algorithm; the mixed loading and unloading scheduling optimization model integrates a model objective function, a combination of model constraints for limiting the model objective function, port yard inventory information, loading and unloading process information, loading and unloading equipment status, and planned port arrival ship information; the model objective function is used to reflect the shortest stay time of all ships at the coal port within the scheduling cycle, and the coal port stay time includes the waiting time for berthing after the ship arrives at the anchorage, berth operation preparation time, loading and unloading operation time, and auxiliary operation time; the improved genetic algorithm is a genetic algorithm improved by a multi-dimensional vector encoding mechanism, population initialization based on process constraint rules, and a fitness function that calls a conflict penalty mechanism;

[0009] Step S2, solving the loading and unloading mixed scheduling optimization model based on the improved genetic algorithm to obtain an optimal scheduling plan; the optimal scheduling plan includes an optimal ship schedule, stacking allocation rules, loading and unloading process arrangement rules, and berth allocation rules; the ship schedule includes the docking time, the start time of operation, and the departure time; the stacking allocation rules include the allocation of stacking spaces in the yard corresponding to the stacker-reclaimer for each ship; the loading and unloading process arrangement rules include the loading and unloading process and equipment usage of each ship; and the berth allocation rules include the allocation of berths for ship operations;

[0010] In step S1, the model constraint combination includes:

[0011] A first model constraint condition, used to constrain the model objective function so as to set a plurality of first restriction rules for determining unique operating conditions for the ship operation process;

[0012] A second model constraint condition is used to constrain the model objective function so as to set a plurality of second restriction rules for the stacking site resource status parameters;

[0013] a third model constraint condition, used to constrain the model objective function so as to set a third restriction rule for berth selection of a ship;

[0014] The fourth model constraint condition is used to constrain the model objective function so as to set at least two fourth restriction rules for the coal port loading and unloading process flow.

[0015] Furthermore, the model objective function is:

[0016]

[0017] In formula (1), V is the set of all ships arriving at the port during the scheduling period, is the actual arrival time of vessel v at the port, The time when the operation of vessel v is completed and leaves the port.

[0018] Furthermore, the first restriction rules are:

[0019] Each arriving vessel must complete all operations at the assigned berth in one go, and a berth can serve at most one vessel at any one time;

[0020] Vessels waiting to be unloaded are only allowed to berth at the unloading berth, and vessels waiting to be loaded are only allowed to complete operations at the loading berth;

[0021] Each vessel's operation must be configured with fixed loading and unloading equipment and process flow, and the loading and unloading equipment or process flow cannot be changed during the operation;

[0022] The second restriction rules are:

[0023] When unloading, the target stack must match the coal type and the remaining capacity must not be less than the ship's load capacity;

[0024] When loading, the current inventory of the stacking position must meet the ship's demand;

[0025] The sum of the number of idle yard scales and the number of occupied yard scales at a specified time t must not be greater than the total number of scales in the corresponding division;

[0026] The third restriction rule is that each vessel is only allowed to berth at an idle berth;

[0027] The two fourth restriction rules are:

[0028] The loading and unloading equipment and the associated yard processes must be in an idle state and have matching processes;

[0029] At the same time, the same stacker-reclaimer is not allowed to be used for both unloading and loading operations.

[0030] Furthermore, the time when the ship v completes its operation and leaves the port is The formula is:

[0031]

[0032] In formula (2), is the operation start time of vessel v, is the auxiliary operation time of ship v, t v is the duration of loading and unloading operations of vessel v, Q v is the loading and unloading volume, and E is the efficiency of the distribution equipment.

[0033] Furthermore, the first model constraint condition is expressed by formula (3) to formula (6), which are respectively:

[0034]

[0035] The second model constraint is expressed by formula (7)-formula (9), which are:

[0036]

[0037] The third model constraint is expressed by formula (10), which is:

[0038]

[0039] The fourth model constraint is expressed by formula (11)-formula (13), which are:

[0040]

[0041] k x h ps +k z h pw ≠2 (13)

[0042] In formula (3)-formula (13), V x represents the set of ships to be unloaded within the scheduling period, V z represents the set of ships to be loaded within the scheduling period, B represents the set of port berths, and B x represents the set of port unloading berths, B z represents the set of port loading berths, C represents the set of port unloaders, D represents the set of port loaders, SR represents the set of port stackers, M represents the set of coal types that can be stored in the yard, and M represents the set ofx Represents the set of coal types to be unloaded, M z Represents the set of coal types to be loaded, P x represents the set of all unloading processes at the port, P z represents the set of all loading processes in the port, Y represents the set of port yards, S represents the set of free yard stacks, W represents the set of occupied yard stacks, G x Indicates the amount of coal to be unloaded, G z Indicates the amount of coal to be loaded, Q v represents the loading and unloading volume of ship v, E c represents the operating efficiency of the ship unloader c, E d represents the operating efficiency of the ship loader d, N y Indicates the number of scales divided by the yard y, N s Indicates the number of idle scales in the yard at time t, N w Indicates the number of scales occupied by the yard at time t, G y Indicates the maximum stockpile volume at each scale of the yard y, f xb ∈{0,1} indicates whether the vessel v to be unloaded x Is it allocated to unloading berth B? x , f zb ∈{0,1} indicates whether the ship v to be loaded z Is it allocated to loading berth B? z , r px ∈{0,1} indicates whether the unloader c is assigned to the unloading process p x , r pz ∈{0,1} indicates whether the loader d is assigned to the loading process p z , h ps ∈{0,1} indicates whether the unloading process p x Stack positions s, h assigned to yard y pw ∈{0,1} indicates whether the shipping process p z Stack w, k allocated to yard y x ∈{0,1} indicates whether the stacker-reclaimer sr is assigned to the unloading process p x , k z ∈{0,1} indicates whether the stacker-reclaimer sr is assigned to the loading process p z .

[0043] Furthermore, in step S2, the execution process of the improved genetic algorithm includes steps S21 to S24 executed sequentially:

[0044] Step S21, performing gene encoding; in step S21, an initial chromosome structure is generated according to the time sequence of ship arrival, and a multi-dimensional vector encoding mechanism is used to represent the distribution relationship of resources including ships, berths, equipment, and stacking positions;

[0045] Step S22, initializing the population;

[0046] Step S23, performing fitness evaluation; in step S23, calculating individual fitness by calling the fitness function of the conflict penalty mechanism;

[0047] Step S24, perform iterative optimization; in step S24, check the termination conditions of the generated population. When the termination conditions are met, perform optimal decoding, expand the chromosome gene sequence, and obtain the best scheduling plan; when the termination conditions are not met, that is, the convergence standard is not reached, continue to perform genetic operations, generate a new population, and return to step S23.

[0048] Furthermore, the execution process of the multi-dimensional vector encoding mechanism includes the following steps S211 to S213 executed in sequence:

[0049] Step S211: compressing the coding dimension according to the corresponding coupling relationship between decision variables in the actual production of the coal terminal;

[0050] Step S212, compressing and simplifying the decision variable combination including ship arrival sequence, ship departure sequence, berth allocation, yard stack allocation, operation process allocation, ship loader selection and stacker-reclaimer selection;

[0051] Step S213: compress the original multi-dimensional decision variables into three decision variables: ship, berth, and stacker / reclaimer, and perform genetic encoding. This effectively reduces the algorithm computational complexity while maintaining the integrity of the scheduling process.

[0052] In step S22, three constraint rules are set for the initialization of the population, and an initial feasible solution is generated in combination with Monte Carlo sampling to ensure the quality and effectiveness of the initial population. The three constraint rules are the first constraint rule, the second constraint rule, and the third constraint rule. The first constraint rule stipulates that in the berth allocation process of the coal port, the corresponding loading and unloading berths must be allocated according to the order of ship arrival and loading and unloading categories. The second constraint rule stipulates that in the stacking position allocation process of the coal port, the available stacking positions must be allocated according to the type of coal and the loading and unloading volume. The third constraint rule stipulates that in the stacking and reclaiming machine allocation process of the coal port, the corresponding stacking and reclaiming machine must be allocated according to the accessibility conditions of the stacking positions in the yard.

[0053] Step S22 includes steps S221 to S227 performed in the following order:

[0054] Step S221: sorting the ships according to their arrival time, so as to ensure that the service order is consistent with the actual arrival order of the ships;

[0055] Step S222, servicing the ships one by one according to their arrival time sequence to ensure that each ship is served;

[0056] Step S223, allocating berth, stacking, and stacker-reclaimer resources to the corresponding ship;

[0057] Step S224, determining whether the ship has no available berth, yard stacking space or stacker / reclaimer. If so, the ship enters a waiting state. If not, step S225 is executed.

[0058] Step S225: Generate a process flow that satisfies the loading and unloading operations based on the matching of the berth and the stacker / reclaimer. This process flow uses Monte Carlo random sampling combined with a feasible solution space constraint to ensure that the process flow satisfies both temporal and spatial constraints.

[0059] Step S226, updating the status of the berth, stacker-reclaimer, and stacking space, and marking the status as occupied;

[0060] Step S227, check the population size to determine whether the current population size meets the preset requirement. If it is, stop the calculation. If it is not, return to step S222 and continue to generate new individuals.

[0061] In step S23, the fitness function is used to quantify the optimization target into a computable numerical indicator and measure the quality of individual chromosomes; the fitness function is set to be inversely proportional to the total time the ship is in port; a stacker-reclaimer conflict penalty mechanism is called to optimize the fitness function; the stacker-reclaimer conflict penalty mechanism is set to: the greater the number of stacker-reclaimer conflicts, the greater the corresponding penalty value; the stacker-reclaimer conflict is set to: if at a certain time point, multiple tasks use the same stacker-reclaimer at the same time; the stacker-reclaimer conflict penalty value P in the reclaimer conflict penalty mechanism is conflict for:

[0062]

[0063] In formula (14), n is the number of conflicting stackers; C j is the number of conflicts of stacker-reclaimer j; λ is the penalty coefficient of conflict, which is a positive integer, so that the more conflicts there are, the greater the penalty;

[0064] The fitness function is:

[0065]

[0066] In formula (15), P conflict is the penalty value for stacker-reclaimer conflict, is the actual arrival time of vessel v at the port, The time when the operation of vessel v is completed and leaves the port.

[0067] The present invention further proposes a coal port loading and unloading production intelligent scheduling implementation system, which is applied to a coal port including at least two ship loaders, at least two ship unloaders, at least two stackers, at least two loading berths, and at least two unloading berths. The implementation system includes a loading and unloading hybrid scheduling optimization model and an improved genetic algorithm:

[0068] The loading and unloading hybrid scheduling optimization model integrates a model objective function, a combination of model constraints for limiting the model objective function, port yard inventory information, loading and unloading process information, loading and unloading equipment status, and information on ships scheduled to arrive at the port; the model objective function is used to reflect the shortest stay time of all ships at the coal port within the scheduling period, and the stay time at the coal port includes the waiting time for berthing after the ship arrives at the anchorage, the berth operation preparation time, the loading and unloading operation time, and the auxiliary operation time;

[0069] The improved genetic algorithm is a genetic algorithm improved by a multi-dimensional vector encoding mechanism, population initialization based on process constraint rules, and a fitness function that calls a conflict penalty mechanism. It is used to solve the loading and unloading mixed scheduling optimization model to obtain the optimal scheduling plan; the optimal scheduling plan includes the optimal ship schedule, stacking allocation rules, loading and unloading process arrangement rules, and berth allocation rules; the ship schedule includes the docking time, the start time of operation, and the departure time; the stacking allocation rules include the allocation of stacking spaces in the yard corresponding to the stacker-reclaimer for each ship; the loading and unloading process arrangement rules include the loading and unloading process and equipment usage of each ship; and the berth allocation rules include the allocation of berths for ship operations;

[0070] The model constraint combination includes:

[0071] A first model constraint condition, used to constrain the model objective function so as to set a plurality of first restriction rules for determining unique operating conditions for the ship operation process;

[0072] A second model constraint condition is used to constrain the model objective function so as to set a plurality of second restriction rules for the stacking site resource status parameters;

[0073] a third model constraint condition, used to constrain the model objective function so as to set a third restriction rule for berth selection of a ship;

[0074] The fourth model constraint condition is used to constrain the model objective function so as to set at least two fourth restriction rules for the coal port loading and unloading process flow.

[0075] The present invention further proposes an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable by the processor. When the processor executes the computer program, the steps of the method for intelligent scheduling of coal port loading and unloading production as described in any of the above items are implemented.

[0076] The present invention further proposes a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the steps of the method for realizing intelligent scheduling of coal port loading and unloading production as described in any of the above items.

[0077] The beneficial effects of the present invention are:

[0078] The present invention can realize intelligent scheduling of mixed loading and unloading production at the southern transit coal terminal, obtain the optimal scheduling plan, and thus maximize the overall operating efficiency of the coal terminal; the present invention simultaneously improves the scheduling efficiency of the coal port and the efficiency of coal loading and unloading and resource allocation, and can adapt to the scheduling optimization of the southern water-to-water transit type coal port, and is applicable to both loading and unloading operations. By establishing a mixed loading and unloading scheduling optimization model suitable for coal port scheduling and solving it through an improved genetic algorithm, the optimal scheduling plan for the overall transit coal port operation is obtained, thereby improving the overall operating efficiency of the coal port and providing technical support for the intelligent scheduling of the water-to-water transit coal terminal, so that the time all ships to be loaded and unloaded are in port is minimized within the operation plan cycle, thereby improving the operating efficiency and service level of the coal port; the present invention realizes the optimization of coal port resource allocation, reasonably allocates berths, loading and unloading equipment, operation processes, stacking spaces and other resources to each ship, can provide assistance for the formulation of coal port production plans, realize intelligent decision-making for scheduling, and is of great value in improving the production and operation efficiency of the coal port. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] Figure 1 This is a workflow diagram of the method for realizing intelligent scheduling of coal port loading and unloading production according to the present invention;

[0080] Figure 2 A chromosome decoding flow chart based on a berth and a stacker-reclaimer according to the present invention;

[0081] Figure 3 A schematic diagram of a crossover operation of an improved genetic algorithm according to the present invention;

[0082] Figure 4 A plan view of the Xiji Coal Terminal at Guangzhou Port, to which the present invention is applied;

[0083] Figure 5 A diagram showing the corresponding relationship between the loading and unloading process flow and loading and unloading equipment at the Xiji Coal Terminal of Guangzhou Port, to which the present invention is applied;

[0084] Figure 6 This is an optimization curve diagram of the total ship time at Xiji Coal Terminal of Guangzhou Port to which the present invention is applied;

[0085] Figure 7 A Gantt chart for scheduling ship loading and unloading operations at Xiji Coal Terminal in Guangzhou Port to which the present invention is applied;

[0086] Figure 8 is the convergence curve of the traditional genetic algorithm;

[0087] Figure 9 This is the convergence curve of the improved genetic algorithm involved in the present invention.

[0088] Figure 10 The workflow diagram for production scheduling of an existing coal port. DETAILED DESCRIPTION

[0089] To make the purpose, technical solutions and advantages of the present invention clearer, the technical solutions of the present invention will be further clearly and completely described below in conjunction with the embodiments of the present invention. It should be noted that the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0090] It should be understood that the directions or positional relationships indicated by terms such as "up", "down", "front", "back", "left", and "right" are based on the directions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific direction, be constructed and operated in a specific direction. Therefore, they should not be understood as limiting the present invention.

[0091] Terms such as "first," "second," "third," and "fourth" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the quantity of the technical features indicated. Therefore, the definition of "first," "second," "third," and "fourth" may explicitly or implicitly include one or more of such features.

[0092] Example 1

[0093] like Figure 1 As shown:

[0094] This embodiment provides a method for implementing intelligent scheduling of coal port loading and unloading production, which is applied to a coal port including at least two ship loaders, at least two ship unloaders, at least two stackers, at least two loading berths, and at least two unloading berths. The method includes the following steps S1 to S2, which are performed in the following order:

[0095] Step S1, setting a mixed loading and unloading scheduling optimization model and an improved genetic algorithm; the mixed loading and unloading scheduling optimization model integrates a model objective function, a combination of model constraints for limiting the model objective function, port yard inventory information, loading and unloading process information, loading and unloading equipment status, and planned port arrival ship information; the model objective function is used to reflect the shortest stay time of all ships at the coal port within the scheduling cycle, and the coal port stay time includes the waiting time for berthing after the ship arrives at the anchorage, berth operation preparation time, loading and unloading operation time, and auxiliary operation time; the improved genetic algorithm is a genetic algorithm improved by a multi-dimensional vector encoding mechanism, population initialization based on process constraint rules, and a fitness function that calls a conflict penalty mechanism;

[0096] Step S2, solving the loading and unloading mixed scheduling optimization model based on the improved genetic algorithm to obtain an optimal scheduling plan; the optimal scheduling plan includes an optimal ship schedule, stacking allocation rules, loading and unloading process arrangement rules, and berth allocation rules; the ship schedule includes the docking time, the start time of operation, and the departure time; the stacking allocation rules include the allocation of stacking spaces in the yard corresponding to the stacker-reclaimer for each ship; the loading and unloading process arrangement rules include the loading and unloading process and equipment usage of each ship; and the berth allocation rules include the allocation of berths for ship operations;

[0097] In step S1, the model constraint combination includes:

[0098] A first model constraint condition, used to constrain the model objective function so as to set a plurality of first restriction rules for determining unique operating conditions for the ship operation process;

[0099] A second model constraint condition is used to constrain the model objective function so as to set a plurality of second restriction rules for the stacking site resource status parameters;

[0100] a third model constraint condition, used to constrain the model objective function so as to set a third restriction rule for berth selection of a ship;

[0101] The fourth model constraint condition is used to constrain the model objective function so as to set at least two fourth restriction rules for the coal port loading and unloading process flow.

[0102] Specifically, the present invention constructs a mixed loading and unloading scheduling optimization model suitable for coal port scheduling, namely a coal port loading and unloading collaborative scheduling optimization model, which realizes the overall operation efficiency optimization and avoids local optimality. It also improves the traditional genetic algorithm to make it more suitable for solving coal port scheduling optimization.

[0103] Optimally, the model objective function is:

[0104]

[0105] In formula (1), V is the set of all ships arriving at the port during the scheduling period, is the actual arrival time of vessel v at the port, The time when the operation of vessel v is completed and leaves the port.

[0106] Optimally, the multiple first restriction rules are:

[0107] Each arriving ship must complete all operations at the assigned berth in one go, and a berth can serve at most one ship at any one time;

[0108] Vessels waiting to be unloaded are only allowed to berth at the unloading berth, and vessels waiting to be loaded are only allowed to complete operations at the loading berth;

[0109] Each vessel's operation must be configured with fixed loading and unloading equipment and process flow, and the loading and unloading equipment or process flow cannot be changed during the operation;

[0110] The second restriction rules are:

[0111] When unloading, the target stack must match the coal type and the remaining capacity must not be less than the ship's load capacity;

[0112] When loading, the current inventory of the stacking position must meet the ship's demand;

[0113] The sum of the number of idle yard scales and the number of occupied yard scales at a specified time t must not be greater than the total number of scales in the corresponding division;

[0114] The third restriction rule is that each vessel is only allowed to berth at an idle berth;

[0115] The two fourth restriction rules are:

[0116] The loading and unloading equipment and the associated yard processes must be in an idle state and have matching processes;

[0117] At the same time, the same stacker-reclaimer is not allowed to be used for both unloading and loading operations.

[0118] Optimally, the time when vessel v completes its operation and leaves the port is The formula is:

[0119]

[0120] In formula (2), is the operation start time of vessel v, is the auxiliary operation time of the ship v (preferably, the auxiliary operation time of the ship v includes the berthing operation time, the unberthing operation time, etc., which are fixed values), t v is the duration of loading and unloading operations of vessel v, Q v is the loading and unloading volume, and E is the efficiency of the distribution equipment.

[0121] Optimally, the first model constraint is expressed by formula (3) to formula (6), which are respectively:

[0122]

[0123] The second model constraint is expressed by formula (7)-formula (9), which are:

[0124]

[0125] The third model constraint is expressed by formula (10), which is:

[0126]

[0127] The fourth model constraint is expressed by formula (11)-formula (13), which are:

[0128]

[0129] k x h ps +k z h pw ≠2 (13)

[0130] In formula (3)-formula (13), V x represents the set of ships to be unloaded within the scheduling period, V z represents the set of ships to be loaded within the scheduling period, B represents the set of port berths, and B x represents the set of port unloading berths, B z represents the set of port loading berths, C represents the set of port unloaders, D represents the set of port loaders, SR represents the set of port stackers, M represents the set of coal types that can be stored in the yard, and M represents the set of x Represents the set of coal types to be unloaded, M z Represents the set of coal types to be loaded, P x represents the set of all unloading processes at the port, P z represents the set of all loading processes in the port, Y represents the set of port yards, S represents the set of free yard stacks, W represents the set of occupied yard stacks, G x Indicates the amount of coal to be unloaded, G z Indicates the amount of coal to be loaded, Q v represents the loading and unloading volume of ship v, E c represents the operating efficiency of the ship unloader c, E drepresents the operating efficiency of the ship loader d, N y Indicates the number of scales divided by the yard y, N s Indicates the number of idle scales in the yard at time t, N w Indicates the number of scales occupied by the yard at time t, G y Indicates the maximum stockpile volume at each scale of the yard y, f xb ∈{0,1} indicates whether the vessel v to be unloaded x Is it allocated to unloading berth B? x , f zb ∈{0,1} indicates whether the ship v to be loaded z Is it allocated to loading berth B? z , r px ∈{0,1} indicates whether the unloader c is assigned to the unloading process p x , r pz ∈{0,1} indicates whether the loader d is assigned to the loading process p z , h ps ∈{0,1} indicates whether the unloading process p x Stack positions s, h assigned to yard y pw ∈{0,1} indicates whether the shipping process p z Stack w, k allocated to yard y x ∈{0,1} indicates whether the stacker-reclaimer sr is assigned to the unloading process p x , k z ∈{0,1} indicates whether the stacker-reclaimer sr is assigned to the loading process p z .

[0131] Optimally, in step S2, the execution process of the improved genetic algorithm includes steps S21 to S24 executed sequentially:

[0132] Step S21, performing gene encoding; in step S21, an initial chromosome structure is generated according to the time sequence of ship arrival, and a multi-dimensional vector encoding mechanism is used to represent the distribution relationship of resources including ships, berths, equipment, and stacking positions;

[0133] Step S22, initializing the population;

[0134] Step S23, performing fitness evaluation; in step S23, calculating individual fitness by calling the fitness function of the conflict penalty mechanism;

[0135] Step S24, perform iterative optimization; in step S24, check the termination conditions of the generated population. When the termination conditions are met, perform optimal decoding, expand the chromosome gene sequence, and obtain the best scheduling plan; when the termination conditions are not met, that is, the convergence standard is not reached, continue to perform genetic operations, generate a new population, and return to step S23.

[0136] Further optimized, the iterative optimization in step S24 comprises selection, crossover, and mutation operations within a genetic algorithm. The selection operation uses a roulette wheel selection method to select the next generation of operators, with the probability of selection determined by the individual's fitness. Individuals with higher calculated fitness values ​​have a greater probability of being selected. The crossover operation uses a two-point crossover method to exchange berth and equipment genes. The mutation operation uses an exchange mutation method to randomly select two ships' berths or stackers for exchange. After the crossover and mutation operations, the ship's operating order remains unchanged, but berth and stacker conflicts may occur. The chromosomes after crossover and mutation need to be inspected and corrected to ensure that each ship is assigned to an available berth and stacker, avoiding duplicate resource allocation.

[0137] Specifically, in step S22, during the population initialization process, by serving ships in sequence, screening resources, generating process flows that conform to the operation sequence, and updating resource status, it can be ensured that the initial population meets the operation process constraints and enhance the quality of the initial solution set; in addition, by combining the Monte Carlo random sampling method with the limitation of the feasible solution space, it is ensured that the diversity of the generated solutions can meet the needs of actual production scenarios.

[0138] Optimally, in step S21, the execution process of the multi-dimensional vector encoding mechanism includes the following steps S211 to S213 executed in sequence:

[0139] Step S211: compressing the coding dimension according to the corresponding coupling relationship between decision variables in the actual production of the coal terminal;

[0140] Step S212, compressing and simplifying the decision variable combination including ship arrival sequence, ship departure sequence, berth allocation, yard stack allocation, operation process allocation, ship loader selection and stacker-reclaimer selection;

[0141] In step S213, the original multi-dimensional decision variables are compressed into three decision variables of ship, berth, and stacker / reclaimer for genetic encoding, thereby effectively reducing the computational complexity of the algorithm while maintaining the integrity of the scheduling process.

[0142] Optimally, in step S22, three constraint rules are set for the initialization of the population, and the initial feasible solution is generated in combination with Monte Carlo sampling, so as to ensure the quality and effectiveness of the initial population; the three constraint rules are the first constraint rule, the second constraint rule, and the third constraint rule; the first constraint rule is that in the berth allocation process of the coal port, the corresponding loading and unloading berths must be allocated according to the order of arrival of ships and the loading and unloading categories; the second constraint rule is that in the yard stack allocation process of the coal port, the available stacks must be allocated according to the type of coal and the loading and unloading volume; the third constraint rule is that in the stacker-reclaimer allocation process of the coal port, the corresponding stacker-reclaimer must be allocated according to the accessibility conditions of the yard stacks.

[0143] Optimally, step S22 includes steps S221 to S227 performed in the following order:

[0144] Step S221: sorting the ships according to their arrival time, so as to ensure that the service order is consistent with the actual arrival order of the ships;

[0145] Step S222, servicing the ships one by one according to their arrival time sequence to ensure that each ship is served;

[0146] Step S223 allocates berth, stack, and stacker / reclaimer resources to the corresponding ship. Referring to a preferred specific implementation process, eligible resources are selected from available berths, stacks, and stackers based on a combination of conditions including ship type, loading capacity, and coal type. For example, a loop structure is used to traverse each element in an array and call the select_resources function to allocate available resources to each element. The select_resources function is typically used in resource allocation or optimization scenarios. Its core function is to dynamically select or match available resources based on the characteristics of the input object.

[0147] Step S224, determining whether the ship has no available berth, yard stacking space or stacker / reclaimer. If so, the ship enters a waiting state. If not, step S225 is executed.

[0148] In step S225, a process flow that satisfies the loading and unloading operations is generated based on the matching of the berth and the stacker-reclaimer. Monte Carlo random sampling is combined with a feasible solution space constraint method to ensure that the process flow satisfies the timing and space constraints. Referring to the preferred specific implementation process, a typical process flow generation function generate_process_flow() is used to generate a process flow that satisfies the loading and unloading operations.

[0149] Step S226, updating the status of the berth, stacker-reclaimer, and yard stack, and marking the status as occupied; referring to the preferred specific implementation process, the update is implemented through the UpdateResource function;

[0150] Step S227, check the population size and determine whether the current population size meets the preset requirements. If it is determined to be yes, stop the calculation; if it is determined not to be, return to step S222 and continue to generate new individuals.

[0151] Optimally, in step S23, the fitness function is used to quantify the optimization target into a computable numerical indicator and measure the quality of individual chromosomes; the fitness function is set to be inversely proportional to the total time the ship is in port; a stacker-reclaimer conflict penalty mechanism is called to optimize the fitness function; the stacker-reclaimer conflict penalty mechanism is set to: if the number of stacker-reclaimer conflicts is greater, the corresponding penalty value is greater; the stacker-reclaimer conflict is set to: if at a certain time point, multiple tasks use the same stacker-reclaimer at the same time; the stacker-reclaimer conflict penalty value P in the reclaimer conflict penalty mechanism is conflict for:

[0152]

[0153] In formula (14), n is the number of conflicting stackers; C j is the number of conflicts of stacker-reclaimer j; λ is the penalty coefficient of conflict, which is a positive integer, so that the more conflicts there are, the greater the penalty;

[0154] The fitness function is:

[0155]

[0156] In formula (15), P conflict is the penalty value for stacker-reclaimer conflict, is the actual arrival time of vessel v at the port, The time when the operation of vessel v is completed and leaves the port.

[0157] Specifically, the stacker-reclaimer is a bucket wheel stacker-reclaimer.

[0158] In this way, the present invention analyzes the correspondence between variables in the encoding process, performs dimensionality compression processing on decision variables, and reduces the difficulty of calculation; the present invention also constructs a population initialization model based on process constraints, and adopts a method that combines Monte Carlo random sampling with feasible solution space limitation to improve the quality of the initial population and the calculation efficiency; the present invention also fully considers the problem of stacker call conflicts during loading and unloading operations at the fitness function design level, introduces a fitness function with a conflict penalty mechanism, and achieves an effective balance between the algorithm convergence speed and constraint violation suppression by adjusting the penalty coefficient.

[0159] Specifically, in the improved genetic algorithm of this embodiment, the fitness function is the core evaluation mechanism of the optimization process. Its essence is to quantify the optimization goal into a computable numerical indicator, which is used to measure the quality of individual chromosomes (i.e., a specific scheduling solution). The design of the fitness function directly affects the search direction and convergence efficiency of the algorithm. When solving the technical problem of production scheduling for loading and unloading at coal terminals, the fitness evaluation value of an individual is negatively correlated with its objective function value. That is, the better the degree of optimization of the objective function value (the shorter the total time the ship is in port), the higher the individual's fitness value, and the easier it is to be selected. Therefore, the technical solution of the present invention designs the fitness function to be inversely proportional to the sum of the ship's time in port.

[0160] In addition, in the production scheduling process of the coal terminal, the resource allocation conflict problem also needs to be considered. For example, referring to the model constraint formula (13), the stacker-reclaimer is a yard loading and unloading equipment that can participate in both the unloading and loading operations. As a key resource, the scheduling conflict of the stacker-reclaimer may affect the efficiency of the overall scheduling. In order to improve the algorithm's screening of population quality and improve the optimization efficiency, the technical solution of the present invention designs a fitness function for the stacker-reclaimer call conflict penalty mechanism.

[0161] In this embodiment, the technical solution of the present invention focuses on the core difficulty of resource allocation in the loading and unloading links in order to solve the optimization problem of the existing coal port multi-link collaborative scheduling system, and designs a corresponding solution algorithm; this optimization problem is essentially an NP-Hard type complex system decision problem in the field of combinatorial optimization. The mathematical model involved in the technical solution of the present invention integrates multi-dimensional decision variables, including time and space dimensions (ship arrival and departure sequence), logistics attribute dimensions (ship type and load, cargo type, etc.), process constraint dimensions (equipment linkage logic, berth equipment stacking correspondence), etc.; given that the solution space of this optimization problem presents a geometric growth characteristic, the computational complexity of the traditional precise algorithm is exponential; therefore, the technical solution of the present invention, based on the construction of a mixed integer programming model, innovatively designs an improved genetic algorithm solution framework to improve the algorithm solution efficiency.

[0162] In addition, traditional genetic algorithms all use a random method to generate the initial population. Although this method can ensure population diversity, it has obvious limitations when dealing with the multi-resource coupling scheduling problem of coal terminals with strong constraints. For the technical problem of coal terminal production scheduling to be solved by the present invention, the berth assigned to each ship after arriving at the port corresponds to the loading and unloading equipment, and the loading and unloading equipment also corresponds to the process and the stacking position in the yard. This results in a large number of randomly generated initial populations being invalid. The large-scale existence of such invalid solutions not only wastes computing resources, but also causes oscillation and divergence in the algorithm convergence trajectory. In this embodiment, in order to ensure the quality of the initial population, reduce the solution space, and improve the algorithm efficiency, the technical solution of the present invention designs an improved genetic algorithm based on the actual situation of on-site production scheduling at the coal terminal. On the basis of the traditional genetic algorithm, the process rule constraints are integrated to generate the initial individuals to ensure that each generated individual (solution) is valid. At the same time, a fitness function that considers the conflict penalty mechanism of the stacker-reclaimer call is designed to achieve an effective balance between the algorithm convergence speed and the constraint violation suppression, further improving the computational dimension and efficiency of the solution set.

[0163] The present invention fully analyzes the influencing factors of coal terminal production scheduling, determines constraints from four aspects: operation process, yard, ship berthing, and loading and unloading process, and takes ships, berths, unloaders, loaders, yard processes and stacking space allocation as decision variables. A water-to-water transfer coal terminal loading and unloading production collaborative scheduling model with the shortest total ship port time as the optimization goal is established, and a genetic algorithm based on process constraints is designed to solve the model; the present invention simplifies the decision variables in the encoding process, uses decision variable dimensionality compression to implement dimensionality reduction processing on the problem space, and normalizes the solution vector by establishing a two-dimensional constraint matrix; in terms of population initialization, the present invention constructs a population initialization model based on process constraints, and adopts a method combining Monte Carlo random sampling with feasible solution space limitation to ensure that the initial solution set meets preset boundary conditions such as the operation timing and spatial layout of the loading and unloading equipment, thereby improving the quality of the initial population; at the fitness function design level, the problem of stacker-reclaimer call conflicts during the loading and unloading operation is fully considered, and a fitness function with a conflict penalty mechanism is introduced. By adjusting the penalty coefficient, an effective balance is achieved between the algorithm convergence speed and constraint violation suppression.

[0164] Example 2

[0165] Example 2 is an optimized design of any one of the technical solutions in Example 1;

[0166] like Figure 1 As shown:

[0167] For further optimization, the optimal scheduling plan also involves the time of arrival of the ship at the port, the type of ship arriving at the port, the ship's pre-berthing position, the combination plan of loading and unloading equipment, the operation time window, the allocation of yard stacking space and the determination of the unloading or loading process.

[0168] Further optimized, before step S1, the scheduling method further includes: setting the scheduling of the coal port according to multiple assumptions, and the multiple assumptions are as follows:

[0169] Assumption 1): Vessel operation requirements are clear and unchangeable. In actual operations, port operators can obtain complete forecast information in advance, including key information such as estimated arrival time, coal types to be loaded and unloaded, cargo load capacity, and loading and unloading process technical indicators;

[0170] Assumption 2): Each vessel completes all operations at the berth in one operation. No berthing changes or replacement of loading and unloading equipment occur during the operation. Each vessel can only be served once.

[0171] Assumption 3): Vessels are scheduled for service according to their arrival schedule;

[0172] Assumption 4): The vessel's time in port begins with arrival and ends with completion of loading and unloading operations and departure from the port. This time span includes scheduling waiting time, auxiliary operation time during berthing and unberthing, and actual operation time.

[0173] Assumption 5): Vessel waiting time is divided into two phases: the first is the time it takes for a vessel to arrive at the port and wait for a berth to be allocated; the second is the time it takes for a vessel to berth and wait for equipment to be allocated;

[0174] Assumption 6): Each stockpile can only store or retrieve materials at a time; simultaneous stacking and retrieval operations are not allowed, and only one type of coal can be used at a stockpile at a time. Assumption 7): The efficiency of loading and unloading operations is determined by the performance of dedicated equipment (the unloader determines the unloading rate, and the loader determines the loading rate). A planned maintenance mechanism is implemented to ensure steady-state operation of the equipment, and the operation process will not be interrupted by sudden failures or process path changes.

[0175] In the actual operation of coal ports, the operating efficiency of loading and unloading equipment exhibits dynamic fluctuations. To facilitate the construction and computational analysis of a hybrid loading and unloading scheduling optimization model, the operating efficiency of loading and unloading equipment is assumed to be fixed. This hybrid loading and unloading scheduling optimization model aims to minimize the total time a vessel spends in port. This time is comprised of multiple components, including the waiting time after arriving at the anchorage, berth preparation time, loading and unloading operations, and auxiliary operations. This total time reflects not only the efficiency of loading and unloading machinery but also a comprehensive assessment of operational performance across multiple dimensions, including berth scheduling efficiency, document processing timeliness, and the responsiveness of supporting services. It is a key parameter for evaluating the competitiveness of coal ports' comprehensive services.

[0176] Optimally, the stacking yard resource status parameters include free stacking yard capacity, inventory of stacking yards to be loaded, and scaled capacity of stacking yards.

[0177] Optimally, the third model constraint conditions are specifically set as follows:

[0178] At the same time, there are only two states of berths: free or occupied. Indicates that the corresponding berth is available at time t. It means that the berth is occupied at time t and each ship can only berth at an idle berth.

[0179] Optimally, the fourth model constraint condition is specifically set as follows:

[0180] 1) The compatibility constraints between ships and loading and unloading equipment are required:

[0181] The loading and unloading equipment and the associated yard processes must be in an idle state and have matching processes; since the stacker-reclaimer can be used for both unloading and loading processes, to ensure that the relevant equipment can be used, it is necessary to confirm whether the equipment associated with the process is in an idle state before arranging the process; assuming Indicates time t and berth B x The connected ship unloader is idle, Indicates time t and berth B x The connected ship unloader is occupied; Indicates time t and berth B z The connected ship loader is idle, Indicates time t and berth B z The connected ship loader is occupied; indicates that the process associated with the unloading operation is idle at time t, It means that the process associated with the unloading operation at time t is occupied; Indicates that the process associated with the loading operation is idle at time t, Indicates that the process associated with the loading operation at time t is occupied. When arranging the ship loading and unloading process, the following requirements must be met:

[0182]

[0183] 2) Process conflict constraints are required:

[0184] In the layout of water-to-water transfer terminal projects, unloading berths and loading berths are usually set up independently, but there are correlations and conflicts in the use of yard equipment. In the loading and unloading process, the unloading process and the loading process share loading and unloading equipment. Among them, the stacker-reclaimer can be used for both the unloading process and the loading process due to its integrated stacking and reclaiming function design. The same bucket wheel excavator can only be used for the unloading process or the loading process in the same time period. When arranging loading and unloading operations at the same time, the unloading process is usually given priority. When the unloading conditions are met, the loading process is considered.

[0185] At the same time, the same stacker-reclaimer cannot be used for both unloading and loading operations at the same time:

[0186] k x h ps +k z h pw ≠2.

[0187] Further optimized, in step S211, the production scheduling decision variables of the coal terminal mainly include the ship arrival timing, ship departure timing, berth allocation, yard stack allocation, operation process allocation, loader selection, stacker selection and other decision variables; in the genetic algorithm solution framework, the individual encoding adopts a multi-layer chromosome structure design, and each gene fragment maps a specific scheduling scheme; in the actual production of the coal terminal, there is a multi-dimensional coupling relationship between these decision variables, and the ship's departure time, ship cargo loading capacity, berth selection, loading and unloading equipment selection (operation efficiency) and operation process arrangement are closely related. A stacker-reclaimer can only operate on the stacks on both sides of the track, and the process is relatively fixed. Ship loaders and unloaders also correspond to fixed berths. For example, there is a fixed correspondence between ship loaders and loading berths. As shown in Table 1 below, the berths that can be served by ship loader SL4 are B3, B4, and B5, while the berths that can be served by ship loader SL5 are B4, B5, and B6. Therefore, based on the correspondence between process constraints between decision variables, the technical solution of the present invention proposes a coding dimension compression strategy, which can appropriately simplify the decision variables to reduce the difficulty of solving the problem.

[0188] berth Ship loader B3 SL4 B4 SL4, SL5 B5 SL4, SL5 B6 SL5

[0189] Table 1

[0190] By analyzing the relationships among the aforementioned variables, the scheduling problem of loading and unloading ships can be transformed into a decision-making problem for berth selection and stacker-reclaimer selection. The choice of stacking sites in the yard is determined by the type of coal to be loaded and unloaded by the berthing ship or by customer demand.

[0191] In summary, to facilitate the scheduling of actual production operations at the terminal, the decision variables that play a key role in the scheduling plan are encoded based on the corresponding relationship between the process constraints of each decision variable in actual production. This paper uses a dimensionality reduction method to compress the original multi-dimensional decision variables into three decision variables: ship, berth, and stacker / reclaimer for genetic encoding. This significantly reduces the algorithm's computational complexity while maintaining the completeness of the scheduling plan. The chromosome encoding adopts a hierarchical genetic structure, and the chromosome encoding relationship between the three is shown in Table 2 below:

[0192] Ship serial number V1 V2 V3 V4 ... Vn Berth number B1 B2 B3 B4 ... Bn Stacker-reclaimer number SR1 SR2 SR3 SR4 ... SRn

[0193] Table 2

[0194] The ship serial number in Table 2 indicates the dispatch order of the ships, which are arranged according to the time of arrival of the ships at the port;

[0195] Berth numbers in Table 2: indicate the berth allocation in the corresponding ship scheduling plan;

[0196] The stacker-reclaimer numbers in Table 2 indicate the allocation of stackers to the corresponding ship dispatching plans;

[0197] According to the coding design in Table 2, the scheduling arrangements corresponding to all ships can be represented by the following two-dimensional array (see the following formula (16) for details). The serial number of the array column corresponds to the ship serial number, and each column represents the scheduling arrangement of the corresponding sequence of ships;

[0198]

[0199] Further optimized, the classical genetic algorithm framework generally adopts a probability model based on uniform distribution to randomly generate the initial population. The classical genetic algorithm framework has good performance in dealing with weakly constrained combinatorial optimization problems, but has certain limitations in the multi-resource scheduling scenario of coal terminals with strong correlation characteristics. The technical solution of the present invention addresses the technical problems of production scheduling at coal terminals. There are corresponding relationships such as the matching of ships and berths, the accessibility of equipment and process flow, and the accessibility of stacks and equipment. To ensure the effectiveness of the initial population solution, in step S22, the relevant process constraint rules for the initialization of the population are set, including:

[0200] Rule 1) Berth allocation requires that loading and unloading berths be assigned according to the order in which ships arrive at the port and the loading and unloading categories. For example, in the example studied in this paper, ships to be unloaded are randomly assigned to unloading berths B1 and B2, and barges to be loaded are randomly assigned to loading berths B3, B4, B5, and B6. Yard stacking locations and stackers are also randomly assigned based on their accessibility to the corresponding processes.

[0201] Rule 2) In allocating stacking spaces in the stockpile yard, the available stacking spaces must be allocated according to the type of coal and the amount of coal loaded and unloaded;

[0202] Rule 3) Regarding the allocation of stackers and reclaimers, considering that stackers and reclaimers can only operate on the stacks on both sides of the track, the stackers and reclaimers are allocated according to the accessibility of the stacks in the yard.

[0203] Further optimized, in step S24 of this embodiment, the selection, crossover and mutation operation process of the genetic algorithm is as follows:

[0204] The selection operation of the improved genetic algorithm of this embodiment is described as follows:

[0205] In existing genetic algorithms, operator selection is a very important step in the population evolution process. It determines how to select high-quality individuals from the current population as parents for subsequent crossover and mutation operations. The selection operation follows the "survival of the fittest" principle. By quantifying the fitness value of the individual, a positive correlation is established between fitness and selection probability, thereby preserving excellent genetic characteristics. Common selection strategies include roulette wheel selection (selection probability is distributed according to fitness ratio, and the higher the fitness, the greater the probability of being selected), tournament selection (randomly selecting several individuals to compete, and the best one wins), sorting selection (distributing probability based on fitness ranking rather than absolute value to alleviate premature convergence), etc. The selection operation needs to balance the contradiction between algorithm convergence efficiency and population diversity. Too fast convergence will lead to population homogeneity and reduce global search capabilities, while too slow convergence will lead to insufficient convergence efficiency.

[0206] The selection operation of the improved genetic algorithm of this embodiment adopts the roulette wheel selection method to select the next generation operator. The specific operation steps are as follows:

[0207] Step 1: Calculate the total fitness: Assume that there are n individuals in the population and the fitness of individual i is f i , then the total fitness is:

[0208]

[0209] Step 2: Calculate the selection probability of each individual: the probability of each individual being selected is positively correlated with its fitness, and the probability P of individual i being selected as the parent i for:

[0210]

[0211] Step 3, cumulative probability distribution: cumulative probability C i Represents the cumulative probability of individual i being selected:

[0212]

[0213] Step 4, Random Selection: Generate a random number r∈[0,1] through a uniform distribution random number generator, and find the corresponding individual in the cumulative probability distribution. The selected individual should satisfy C i-1 <r ≤ C i condition;

[0214] Step 5: Complete Selection: Repeat Step 4 until the required number of individuals is selected;

[0215] Regarding the crossover operation of the improved genetic algorithm in this embodiment, the following is an explanation:

[0216] In the existing genetic algorithm, the crossover operation is the key evolutionary mechanism that simulates biological gene recombination in the genetic algorithm. By exchanging chromosome segments of parental individuals to generate new offspring, it promotes population diversity. Its main role is to fuse high-quality gene blocks and balance the global search and local development capabilities. Common crossover strategies include single-point crossover, two-point crossover, uniform crossover, etc. The crossover probability controls the evolutionary intensity and needs to be dynamically adjusted according to the problem complexity;

[0217] For the characteristics of the multi-constraint scheduling problem of coal terminals, this paper selects the "two-point crossover" strategy for the crossover operation. Two-point crossover means randomly selecting two crossover points, dividing the chromosomes of the parents into three parts, and then exchanging the genes between these two points to generate new offspring. The specific operation is as follows. Let P1 and P2 be two parental chromosomes:

[0218]

[0219] Randomly select crossover points k1, k2 (satisfying 1 ≤ k1 < k2 ≤ L) to generate offspring C1, C2:

[0220]

[0221] Two-point crossover realizes the discontinuous inheritance of the dominant segments of the parents through the directional recombination of gene modules, maintains the diversity of the population, avoids premature convergence, and enhances the global search ability of the algorithm;

[0222] The crossover operation is as Figure 3 shown. Here, it is assumed that k1 takes the value of 2 and k2 takes the value of 4. After the crossover operation according to the above method, the operation sequence of the ships remains unchanged, but problems such as berth mismatch or occupation and unreachability of stacker-reclaimer stacking may occur. Therefore, the chromosomes after crossover need to be inspected and corrected to ensure the feasibility of the ship scheduling plan to meet the process constraint requirements. The following are the specific operation steps for chromosome correction:

[0223] Step ①: Take out the ships one by one from the chromosomes after crossover according to the ship arrival order;

[0224] Step ②: Check whether the berth assigned to the ship meets the requirements; if the berth meets the requirements, proceed to Step ③; if the berth does not match or is occupied, jump to Step ④ to reassign the berth.

[0225] Step ③: Check whether the stacker-reclaimer assigned to the ship can meet the loading and unloading requirements; if the stacker-reclaimer is available and the yard capacity meets the requirements, proceed to Step ⑤; if the stacker-reclaimer is not available or the yard capacity is insufficient, jump to Step ④.

[0226] Step ④: Reassign the berth and stacker-reclaimer that meet the process constraints, and update the status of the berth, equipment, yard, etc.

[0227] Step ⑤: Determine whether the last ship has been processed; if so, end; otherwise, jump to Step ① to continue processing the next ship.

[0228] The mutation operation of the improved genetic algorithm in this embodiment is described as follows:

[0229] Common mutation operation methods include single-point mutation, multi-point mutation, and swap mutation; for the characteristics of coal collection and distribution scheduling in coal terminals, the improved genetic algorithm in this embodiment adopts swap mutation, that is, randomly select the berths or stacker-reclaimers of two ships for swapping. The specific operation is as follows:

[0230] Let the chromosome C = [g1, g2,..., g L , randomly select two gene positions i, j (satisfying 1 ≤ i < j ≤ L), and after performing the swap mutation operation, C' = [g1,..., g j ,..., g i ,..., g L .

[0231] The mutation operation of the improved genetic algorithm in this embodiment is as follows:

[0232] Assume that ships V1 and V2 are selected. If they are respectively allocated berths B1 and B3, and stackers SR2 and SR4, then after the mutation operation, their berths and stackers are exchanged to obtain a new allocation plan: the berth allocated to V1 becomes B3, and the stacker allocated to V2 becomes SR4; the berth allocated to V2 becomes B1, and the stacker allocated to V2 becomes SR2. After the above mutation operation, the newly generated chromosome individuals may also fail to meet the process constraints, and problems such as berth mismatch, conflict, and unreachable stacker positions may occur. In order to improve the quality of individuals, it is necessary to perform constraint satisfaction inspection and correction on the individuals after the mutation operation. The inspection and correction process is similar to the crossover operation correction and will not be repeated here. The core of the mutation operation design of the improved genetic algorithm in this embodiment is to ensure that while increasing the diversity of the population, the berth and stacker resource constraints are met, ensuring the diversity and effectiveness of the scheduling plan. Through the mutation operation, combined with an appropriate conflict repair strategy, the search ability of the genetic algorithm can be effectively improved to avoid falling into the local optimal solution.

[0233] Further optimized, in step S24 of this embodiment, the chromosome decoding process of the genetic algorithm is described as follows:

[0234] When constructing the production scheduling model of the coal terminal, it is necessary to consider multiple factors that affect the objective function, such as which berth the ship should dock at, which stacker to use, how to arrange the operation time, which stacker the cargo should be stacked at, how to arrange the loading and unloading process, etc. These factors are interrelated. If all of them are put into the model and calculated together, the complexity will be very high. Therefore, when designing the encoding, the technical solution of the present invention adopts dimensionality compression technology to reduce the difficulty of solving the problem. According to the corresponding relationship between the variables, only three key decision variables, namely ships, berths and stackers, are selected. After the algorithm obtains the optimal solution through optimization calculation, it is also necessary to complete the details of other variables through decoding operations to restore the complete scheduling plan. In this embodiment, according to the berth and stacker information contained in the chromosome, decoding is performed specifically through the following steps. The decoding process is as follows: Figure 2 As shown; the chromosome decoding process of the improved genetic algorithm of this embodiment includes the following steps:

[0235] Step I: Determine the stacking positions for the stacker / reclaimer. Based on each stacker / reclaimer's actual position and operating data, define its dedicated operating area on the yard map. For example, a stacker / reclaimer can only operate on two sides of the stack.

[0236] Step II: Determine the loading and unloading process. Determine the loading and unloading process of the ship based on the accessibility and constraints between the stacker and the berth.

[0237] Step III: Determine the operating and berthing times. Calculate the berthing and operating times for each vessel based on the order of arrival, berth allocation, stacker / reclaimer capacity, and equipment availability. After decoding steps I-III, a specific scheduling plan is generated. This plan includes at least the following elements:

[0238] Element (1), ship schedule: the time of arrival, start of operation and departure of each ship;

[0239] Element (2), stack allocation: allocation of stacking positions for each ship to the stacker / reclaimer;

[0240] Element (3), arrangement of loading and unloading process: the loading and unloading process of each stacker and reclaimer, including the specific operation sequence of loading and unloading of each ship;

[0241] Element (4), berth allocation: berth allocation for ship operations.

[0242] As one of the application cases, specifically, the coal port is a coal port that conducts water-to-water transshipment operations.

[0243] Example 3

[0244] This embodiment proposes a coal port loading and unloading production intelligent scheduling implementation system, which is applied to a coal port including at least two ship loaders, at least two ship unloaders, at least two stackers, at least two loading berths, and at least two unloading berths. The implementation system includes a loading and unloading hybrid scheduling optimization model and an improved genetic algorithm;

[0245] The loading and unloading hybrid scheduling optimization model integrates a model objective function, a combination of model constraints for limiting the model objective function, port yard inventory information, loading and unloading process information, loading and unloading equipment status, and information on ships scheduled to arrive at the port; the model objective function is used to reflect the shortest stay time of all ships at the coal port within the scheduling period, and the stay time at the coal port includes the waiting time for berthing after the ship arrives at the anchorage, the berth operation preparation time, the loading and unloading operation time, and the auxiliary operation time;

[0246] The improved genetic algorithm is a genetic algorithm improved by a multi-dimensional vector encoding mechanism, population initialization based on process constraint rules, and a fitness function that calls a conflict penalty mechanism. It is used to solve the loading and unloading mixed scheduling optimization model to obtain the optimal scheduling plan; the optimal scheduling plan includes the optimal ship schedule, stacking allocation rules, loading and unloading process arrangement rules, and berth allocation rules; the ship schedule includes the docking time, the start time of operation, and the departure time; the stacking allocation rules include the allocation of stacking spaces in the yard corresponding to the stacker-reclaimer for each ship; the loading and unloading process arrangement rules include the loading and unloading process and equipment usage of each ship; and the berth allocation rules include the allocation of berths for ship operations;

[0247] The model constraint combination includes:

[0248] A first model constraint condition, used to constrain the model objective function so as to set a plurality of first restriction rules for determining unique operating conditions for the ship operation process;

[0249] A second model constraint condition is used to constrain the model objective function so as to set a plurality of second restriction rules for the stacking site resource status parameters;

[0250] a third model constraint condition, used to constrain the model objective function so as to set a third restriction rule for berth selection of a ship;

[0251] The fourth model constraint condition is used to constrain the model objective function so as to set at least two fourth restriction rules for the coal port loading and unloading process flow.

[0252] Further optimized, a coal port loading and unloading production intelligent scheduling implementation system of this embodiment executes the implementation method of any one of the technical solutions in Example 1 or Example 2.

[0253] This embodiment further proposes an electronic device, comprising a storage device, a processor, and a computer program stored in the storage device and executable by the processor. When the processor executes the computer program, the steps of the method for realizing intelligent scheduling of coal port loading and unloading production as described in any one of the technical solutions of Example 1 or Example 2 are implemented; specifically, the electronic device is used to control the loading and unloading production scheduling of a coal port.

[0254] This embodiment further proposes a computer-readable storage medium, which stores a computer program, and is characterized in that when the computer program is executed by a processor, it implements the steps of the method for realizing intelligent scheduling of coal port loading and unloading production as described in any one of the technical solutions of Example 1 or Example 2.

[0255] Based on the above beneficial effects of the present invention, the significance of the beneficial effects of the present invention is demonstrated through an engineering example:

[0256] An engineering example of the application of the present invention is the Xiji Coal Terminal of Guangzhou Port, which is summarized as follows:

[0257] Xiji Coal Terminal of Guangzhou Port is located on the northeast bank of the Pearl River, in the Guangzhou Economic and Technological Development Zone, at the confluence of the Pearl River and the Dongjiang River. It enjoys a superior geographical location and convenient transportation, connected to railways, highways and waterways. It is one of the major specialized coal loading and unloading terminals in South China and an important hub for the country's "North Coal to South Transport". Since its commissioning, the terminal has undergone many technical upgrades and renovation and expansion projects. The current Xiji Terminal has two 70,000-ton unloading berths, four 4,000-ton loading berths, and nine strip yards with a maximum designed storage capacity of 900,000 tons and an annual throughput of over 10 million tons. The terminal's production process system is equipped with six bridge-type grab ship unloaders, two ship loaders, eight stackers and reclaimers, two loaders, and more than 50 belt conveyors throughout the terminal, as well as several loaders and other mobile equipment and sewage treatment systems. The terminal layout is shown below. Figure 4 shown.

[0258] The loading and unloading process of Xiji Coal Terminal at Guangzhou Port is as follows:

[0259] The two 70,000-ton unloading berths at Xiji Coal Terminal are equipped with six bridge-type grab ship unloaders, of which SU7, SU8, and SU11 are located at unloading berth B1, and SU9, SU10, and SU12 are located at unloading berth B2. They are also equipped with four belt conveyors for the unloading process, one of which is dedicated to the ship-to-ship loading process. The four loading berths are equipped with two ship loaders, SL4 and SL5. SL4 can operate at loading berths B3, B4, and B5, and SL5 can operate at loading berths B4, B5, and B6. They are also equipped with two loading belt conveyors. The two card loaders are equipped with two loading belt conveyors, which are dedicated to loading trains out of the warehouse. Another two belt conveyors are directly delivered to the nearby power plant. The corresponding relationship between the terminal loading and unloading process and loading and unloading equipment is as follows: Figure 5 As shown, in order to simplify the process in the yard area, the connection process of the transfer belt conveyor is omitted, and only the belt conveyor directly connected to the ship unloader and ship loader is shown. Figure 5 B is the berth, SU is the ship unloader, SR is the bucket wheel crane (stacker / reclaimer), SL is the ship loader, RL is the card loader, and BC is the belt conveyor.

[0260] The storage yards at Guangzhou Port Xiji Coal Terminal are distributed as follows:

[0261] The Xiji Coal Terminal at Guangzhou Port has nine strip yards, numbered 0 to 8. Yard 7 is a temporary yard without an accessible stacker / reclaimer. Yard 8 is a dedicated yard for the adjacent power plant and is equipped with two dedicated stackers / reclaimers, the SR7 and SR8. The remaining yards are general-purpose yards with accessible stackers / reclaimers. The allocation of yards to stackers and reclaimers is shown in Table 3, along with relevant parameters for terminal loading and unloading processes.

[0262]

[0263] Table 3

[0264] The main loading and unloading equipment and related facilities parameters of Guangzhou Port Xiji Coal Terminal are shown in Table 4 below (Table 4 is a table of relevant parameters for terminal loading and unloading processes):

[0265] Serial number name Specifications 1 Ship Unloader (SU) Rated 1250t / h, average 625t / h 2 Ship loader (SL) Rated 1500t / h, average 1050th 3 Stacker-Reclaimer (SR) Stacking 2500t / h, reclaiming 1500t / h 4 Berthing time of ocean-going vessels 45min 5 Departure time of ocean-going vessels 45min 6 Barge berthing time 15min 7 Barge departure time 15min 8 Yard design capacity 900,000 tons 9 Number of grids per yard 50 grids 10 Capacity of each grid in the yard 2000t

[0266] Table 4

[0267] The grid division of the storage yard at Xiji Coal Terminal of Guangzhou Port is a virtual gridding performed for the convenience of calculation in the model of the present invention. Each storage yard is divided into 45 grids, and the storage capacity of each grid is 2000t. During the actual loading and unloading operation of the terminal, factors affecting the operating efficiency such as equipment failure, unloader relocation operation, clearance operation, process conversion, etc. must also be considered, and the average operating efficiency of the equipment cannot reach the rated efficiency. Here, when calculating in the model of the present invention, the average operating efficiency of the unloader is calculated as 50% of the rated efficiency, that is, 625t / h; the efficiency of the loader is calculated as 70% of the rated efficiency, that is, 1050t / h; the unloading efficiency is calculated based on the berth efficiency, that is, three unloaders operate at one unloading berth, and the efficiency is 1875t / h; the loading berth operating efficiency is calculated based on the loader efficiency, that is, 1050t / h; the ship departure time is averaged based on past experience, with a large ship departure of 45min and a barge departure of 15min.

[0268] The types of transport vessels at Guangzhou Port Xiji Coal Terminal are as follows:

[0269] Coal transportation in my country's coastal areas is usually carried out from northern ports to southern ports, usually mainly by ships of 50,000 to 100,000 tons, which can ensure the best transportation economic benefits; considering the navigation conditions of each terminal and the water level depth restrictions of some inland port berths, in the actual transportation process, the actual carrying capacity of ships berthed in South China is mostly in the order of 50,000 to 70,000 tons. The Xiji Coal Terminal of Guangzhou Port is an inland port with two unloading berths with a water depth of 12.5 meters. The ships received and unloaded are mainly of 50,000 tons. The coal received and unloaded mainly comes from northern ports, including Qinhuangdao Port, Caofeidian Port, Huanghua Port, Tianjin Port and other coal terminals. A small amount of foreign trade coal comes from Indonesia and other places; the water depth of the loading berth of Xiji Terminal is about 4 meters, and the barge capacity is between 1,000 and 4,000 tons, which is mainly sent to power plants, steel mills, factories and other factories in surrounding areas to meet local power generation, industrial and other needs.

[0270] The present invention is based on the scheduling optimization analysis of Guangzhou Port Xiji Coal Terminal as follows:

[0271] After applying the technical solution of the present invention, an experimental analysis was conducted on the production plan data of a short-cycle operation at the Xiji Coal Terminal of Guangzhou Port. For example, the ship berthing plan data within the 48-hour period from 0:00 on the 17th to 24:00 on the 18th of a certain month was used for example calculation. The planned ship arrival information is shown in Table 5 below (Table 5 is the planned ship arrival information table), and the initial status information of the yard is shown in Table 6 below (Table 6 is the initial status information table of the yard).

[0272]

[0273]

[0274] Table 5

[0275]

[0276]

[0277] Table 6

[0278] Based on a test experiment, the algorithm parameters of the technical solution of the present invention are set according to the following Table 7 (Table 7 is an algorithm operation parameter table):

[0279] Serial number Parameter name Parameter value 1 Population size 80 2 Maximum number of iterations 500 3 Crossover probability 0.7 4 Mutation probability 0.3

[0280] Table 7

[0281] The algorithm of the present invention is programmed and run using MATLAB software. By executing the program of the present invention for implementing the intelligent scheduling method for coal port loading and unloading production (hereinafter referred to as the program of the present invention), application tests are carried out, and the following experimental results are obtained, as follows:

[0282] According to the relevant information such as the loading and unloading process, equipment parameters and planned arrival ship data of the Xiji Coal Terminal of Guangzhou Port, the operation data of the Xiji Coal Terminal of Guangzhou Port was solved and calculated by executing the program of the present invention. The program of the present invention was run 5 times, and the minimum objective function value after 500 iterations was taken as the final result. Figure 6 The optimization curve of the total time a ship spends in port is shown.

[0283] Through Figure 6 According to the analysis, the total time of ships in port under the initial random scheduling scheme in this test experiment is 301.105 hours. As the population evolves, the program of the present invention enters the convergence stage at the 65th generation, and the total time of ships in port begins to stabilize, reaching 191.934 hours. It is particularly noteworthy that in the 72-192 generation range, the program of the present invention continuously outputs a suboptimal solution of 191.882 hours, indicating that the program of the present invention undergoes deep local search in this stage; finally, the convergence threshold is broken through at the 231st generation, the total time of ships in port remains stable, and the optimal solution is 191.522 hours. The calculation result data of the corresponding optimal solution are shown in Table 8 (Table 8 is a list of optimal solution data):

[0284]

[0285]

[0286]

[0287] Table 8

[0288] According to the optimized scheduling data (as shown in Table 8), the waiting time of all ships at the anchorage was 6.452 hours, the waiting time at the berth was 17.298 hours, the actual operation time was 145.272 hours, the auxiliary operation time for ship undocking was 22.5 hours, and the total time in port was 191.522 hours, which was 36.395% shorter than the total time in port of 301.105 hours without optimization. This fully verifies the feasibility of the technical solution of the present invention and the significance of its beneficial effects.

[0289] Decode the optimal solution in Table 8 and obtain the complete ship operation schedule as shown in Table 9 (Table 9 is the ship production operation schedule). Draw the ship production scheduling Gantt chart according to the optimal solution (see Table 9). Figure 7 shown).

[0290]

[0291]

[0292]

[0293] Table 9

[0294] The data dimensions in Table 9 include various process parameters for each ship's production scheduling operation, including detailed information such as berths, operating machinery and equipment, yard stacks, intermediate operation process flows, and berthing and departure times, which can provide a reference for terminal scheduling decisions and production operation plan formulation. Figure 7 A Gantt chart for ship loading and unloading scheduling clearly displays each vessel's waiting time for berthing, berthing time, waiting time for operation, operation time, and departure time, intuitively reflecting the entire scheduling process of each ship during its stay in port. For example, vessel V13 arrived at 4:00 PM on the 17th, began berthing at berth B6 at 4:30 PM, began loading operations at 5:27 PM, and departed at 7:31 PM. Due to a berth being occupied, vessel V13 waited 30 minutes before a vacant berth became available. After berthing, the loader was occupied and could not immediately begin loading operations. Loading operations could only begin after waiting 42 minutes until the loader became available, totaling 108 minutes. The final calculation shows vessel V13's total time in port was 3 hours and 31 minutes.

[0295] The effectiveness analysis of the technical solution of the present invention is as follows:

[0296] According to Table 5 (Ship Arrival Information Table), the number of arriving ships is taken as the first 10, 20, 30, and 40 respectively. When other parameters remain unchanged, CPLEX and the improved genetic algorithm designed in this paper are used to solve the objective function value, and the solution results and solution efficiency are compared and analyzed. The experimental data are shown in Table 10 (Table 10 is a comparison table of the results of CPLEX and the improved algorithm in this paper for solving different scale examples); At the same time, taking the scale of 40 ships as an example, the traditional genetic algorithm and the improved genetic algorithm of this invention are used respectively to compare the convergence efficiency of the algorithm. Figure 8 、 Figure 9 As shown, through Figure 8 、 Figure 9 The convergence of the two algorithms was compared; the experimental platform computer core configuration was an Intel i5-6500 CPU with a base frequency of 3.20GHz, an operating system of Windows 7 x64 Ultimate, and 8.0GB of DDR4 RAM. In Table 10, the gap (GAP) is the percentage of the CPLEX objective function value minus the objective function value of the proposed algorithm. A gap of less than 5% is considered acceptable.

[0297]

[0298] Table 10

[0299] The experimental data in Table 10 show that compared to the CPLEX solver, the proposed program maintains a high level of accuracy in solving the objective function value, with a relative deviation rate within 4%. In terms of computational efficiency, the proposed program has significant advantages over both traditional solvers and traditional genetic algorithms. In particular, when the ship dispatching scale reaches 40 or more, the solution time is approximately 10 seconds, while CPLEX takes over 480 seconds at the same scale, which no longer meets the timeliness requirements of practical application scenarios. Of particular note, as the scale increases, CPLEX's solution time increases exponentially, and may even become unsolvable, while the proposed program still achieves a stable solution and maintains high efficiency. For the same 40-ship scale, the traditional genetic algorithm converges to 186.75 after 328 iterations, while the improved genetic algorithm of the proposed program converges to 186.75 after only 111 iterations, achieving computational efficiency three times that of the traditional genetic algorithm. This demonstrates that the proposed program's improvements in population initialization and fitness function during algorithm design are effective. Table 10 Figure 8 and Figure 9 The experimental data fully verified the effectiveness and superiority of the program of the present invention in solving the coal terminal scheduling optimization model. This is mainly reflected in the fact that, on the one hand, it can ensure a high level of accuracy of the optimal solution, and on the other hand, it has significant advantages in solution efficiency, which makes the program of the present invention have good engineering application value in complex scheduling scenarios.

[0300] In summary, this experimental test describes the basic conditions of the Xiji Terminal in Guangzhou Port, including the loading and unloading processes, yard layout, equipment parameters, and berthing vessel types. Using the program of the present invention, combined with production data from an engineering example terminal, a scheduling plan for 41 ships scheduled to berth within a 48-hour period was solved. The results show that the optimized scheduling plan shortens the ship's time in port by 36.39% compared to the unoptimized plan, verifying the feasibility of the present invention. Finally, a comparison of the solutions obtained with the program of the present invention using CPLEX and a traditional genetic algorithm confirms the effectiveness and engineering applicability of the present invention.

[0301] The above-described embodiments merely illustrate several implementations of the present invention, and while their descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art would be able to make numerous variations and improvements without departing from the spirit of the present invention, all of which fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be determined by the appended claims.

Claims

1. A method for realizing intelligent scheduling of coal port loading and unloading production, characterized in that: Applicable to a coal port comprising at least two ship loaders, at least two ship unloaders, at least two stackers, at least two ship loading berths, and at least two ship unloading berths, the implementation method comprises the following steps S1 to S2 performed in sequence: Step S1, setting a mixed loading and unloading scheduling optimization model and an improved genetic algorithm; the mixed loading and unloading scheduling optimization model integrates a model objective function, a combination of model constraints for limiting the model objective function, port yard inventory information, loading and unloading process information, loading and unloading equipment status, and planned port arrival ship information; the model objective function is used to reflect the shortest stay time of all ships at the coal port within the scheduling cycle, and the coal port stay time includes the waiting time for berthing after the ship arrives at the anchorage, berth operation preparation time, loading and unloading operation time, and auxiliary operation time; the improved genetic algorithm is a genetic algorithm improved by a multi-dimensional vector encoding mechanism, population initialization based on process constraint rules, and a fitness function that calls a conflict penalty mechanism; Step S2, solving the loading and unloading mixed scheduling optimization model based on the improved genetic algorithm to obtain an optimal scheduling plan; the optimal scheduling plan includes an optimal ship schedule, stacking allocation rules, loading and unloading process arrangement rules, and berth allocation rules; the ship schedule includes the docking time, the start time of operation, and the departure time; the stacking allocation rules include the allocation of stacking spaces in the yard corresponding to the stacker-reclaimer for each ship; the loading and unloading process arrangement rules include the loading and unloading process and equipment usage of each ship; and the berth allocation rules include the allocation of berths for ship operations; In step S1, the model constraint combination includes: A first model constraint condition, used to constrain the model objective function so as to set a plurality of first restriction rules for determining unique operating conditions for the ship operation process; A second model constraint condition is used to constrain the model objective function so as to set a plurality of second restriction rules for the stacking site resource status parameters; a third model constraint condition, used to constrain the model objective function so as to set a third restriction rule for berth selection of a ship; The fourth model constraint condition is used to constrain the model objective function so as to set at least two fourth restriction rules for the coal port loading and unloading process flow.

2. The method for realizing intelligent scheduling of coal port loading and unloading production according to claim 1 is characterized in that: The objective function of the model is: In formula (1), V is the set of all ships arriving at the port during the scheduling period, is the actual arrival time of vessel v at the port, The time when the operation of vessel v is completed and leaves the port.

3. The method for realizing intelligent scheduling of coal port loading and unloading production according to claim 2 is characterized in that: The multiple first restriction rules are: Each arriving vessel must complete all operations at the assigned berth in one go, and a berth can serve at most one vessel at any one time; Vessels waiting to be unloaded are only allowed to berth at the unloading berth, and vessels waiting to be loaded are only allowed to complete operations at the loading berth; Each vessel's operation must be configured with fixed loading and unloading equipment and process flow, and the loading and unloading equipment or process flow cannot be changed during the operation; The second restriction rules are: When unloading, the target stack must match the coal type and the remaining capacity must not be less than the ship's load capacity; When loading, the current inventory of the stacking position must meet the ship's demand; The sum of the number of idle yard scales and the number of occupied yard scales at a specified time t must not be greater than the total number of scales in the corresponding division; The third restriction rule is that each vessel is only allowed to berth at an idle berth; The two fourth restriction rules are: The loading and unloading equipment and the associated yard processes must be in an idle state and have matching processes; At the same time, the same stacker-reclaimer is not allowed to be used for both unloading and loading operations.

4. The method for realizing intelligent scheduling of coal port loading and unloading production according to claim 3 is characterized in that: Time when ship v completes its operation and leaves the port The formula is: In formula (2), is the operation start time of vessel v, is the auxiliary operation time of ship v, t v is the duration of loading and unloading operations of vessel v, Q v is the loading and unloading volume, and E is the efficiency of the distribution equipment.

5. The method for realizing intelligent scheduling of coal port loading and unloading production according to claim 3 or 4, characterized in that: The first model constraint condition is expressed by formula (3) to formula (6), which are respectively: The second model constraint is expressed by formula (7)-formula (9), which are: The third model constraint is expressed by formula (10), which is: The fourth model constraint is expressed by formula (11)-formula (13), which are: k x h ps +k z h pw ≠2 (13) In formula (3)-formula (13), B x represents the set of ships to be unloaded within the scheduling period, B z represents the set of ships to be loaded within the scheduling period, B represents the set of port berths, and B x represents the set of port unloading berths, B z represents the set of port loading berths, C represents the set of port unloaders, D represents the set of port loaders, SR represents the set of port stackers, M represents the set of coal types that can be stored in the yard, and M represents the set of x Represents the set of coal types to be unloaded, M z Represents the set of coal types to be loaded, P x represents the set of all unloading processes at the port, P z represents the set of all loading processes in the port, Y represents the set of port yards, S represents the set of free yard stacks, W represents the set of occupied yard stacks, G x Indicates the amount of coal to be unloaded, G z Indicates the amount of coal to be loaded, Q v represents the loading and unloading volume of ship v, E c represents the operating efficiency of the ship unloader c, E d represents the operating efficiency of the ship loader d, N y Indicates the number of scales divided by the yard y, N s Indicates the number of idle scales in the yard at time t, N w Indicates the number of scales occupied by the yard at time t, G y Indicates the maximum stockpile volume at each scale of the yard y, f xb ∈{0,1} indicates whether the vessel v to be unloaded x Is it allocated to unloading berth B? x , f zb ∈{0,1} indicates whether the ship v to be loaded z Is it allocated to loading berth B? z , r px ∈{0,1} indicates whether the unloader c is assigned to the unloading process p x , r pz ∈{0,1} indicates whether the loader d is assigned to the loading process p z , h ps ∈{0,1} indicates whether the unloading process p x Stack positions s, h assigned to yard y pw ∈{0,1} indicates whether the shipping process p z Stack w, k allocated to yard y x ∈{0,1} indicates whether the stacker-reclaimer sr is assigned to the unloading process p x , k z ∈{0,1} indicates whether the stacker-reclaimer sr is assigned to the loading process p z 。 6. The method for realizing intelligent scheduling of coal port loading and unloading production according to any one of claims 1 to 4, characterized in that: In step S2, the execution process of the improved genetic algorithm includes steps S21 to S24 executed sequentially: Step S21, performing gene encoding; in step S21, an initial chromosome structure is generated according to the time sequence of ship arrival, and a multi-dimensional vector encoding mechanism is used to represent the distribution relationship of resources including ships, berths, equipment, and stacking positions; Step S22, initializing the population; Step S23, performing fitness evaluation; in step S23, calculating individual fitness by calling the fitness function of the conflict penalty mechanism; Step S24, perform iterative optimization; in step S24, check the termination conditions of the generated population. When the termination conditions are met, perform optimal decoding, expand the chromosome gene sequence, and obtain the best scheduling plan; when the termination conditions are not met, that is, the convergence standard is not reached, continue to perform genetic operations, generate a new population, and return to step S23.

7. The method for realizing intelligent scheduling of coal port loading and unloading production according to claim 6 is characterized in that: In step S21, the execution process of the multi-dimensional vector encoding mechanism includes the following steps S211 to S213 executed in sequence: Step S211: compressing the coding dimension according to the corresponding coupling relationship between decision variables in the actual production of the coal terminal; Step S212, compressing and simplifying the decision variable combination including ship arrival sequence, ship departure sequence, berth allocation, yard stack allocation, operation process allocation, ship loader selection and stacker-reclaimer selection; Step S213: compress the original multi-dimensional decision variables into three decision variables: ship, berth, and stacker / reclaimer, and perform genetic encoding. This effectively reduces the algorithm computational complexity while maintaining the integrity of the scheduling process. In step S22, three constraint rules are set for the initialization of the population, and an initial feasible solution is generated in combination with Monte Carlo sampling to ensure the quality and effectiveness of the initial population. The three constraint rules are the first constraint rule, the second constraint rule, and the third constraint rule. The first constraint rule stipulates that in the berth allocation process of the coal port, the corresponding loading and unloading berths must be allocated according to the order of ship arrival and loading and unloading categories. The second constraint rule stipulates that in the stacking position allocation process of the coal port, the available stacking positions must be allocated according to the type of coal and the loading and unloading volume. The third constraint rule stipulates that in the stacking and reclaiming machine allocation process of the coal port, the corresponding stacking and reclaiming machine must be allocated according to the accessibility conditions of the stacking positions in the yard. Step S22 includes steps S221 to S227 performed in the following order: Step S221: sorting the ships according to their arrival time, so as to ensure that the service order is consistent with the actual arrival order of the ships; Step S222, servicing the ships one by one according to their arrival time sequence to ensure that each ship is served; Step S223, allocating berth, stacking, and stacker-reclaimer resources to the corresponding ship; Step S224, determining whether the ship has no available berth, yard stacking space or stacker / reclaimer. If so, the ship enters a waiting state. If not, step S225 is executed. Step S225: Generate a process flow that satisfies the loading and unloading operations based on the matching of the berth and the stacker / reclaimer. This process flow uses Monte Carlo random sampling combined with a feasible solution space constraint to ensure that the process flow satisfies both temporal and spatial constraints. Step S226, updating the status of the berth, stacker-reclaimer, and stacking space, and marking the status as occupied; Step S227, check the population size to determine whether the current population size meets the preset requirement. If it is, stop the calculation. If it is not, return to step S222 and continue to generate new individuals. In step S23, the fitness function is used to quantify the optimization target into a computable numerical indicator and measure the quality of individual chromosomes; the fitness function is set to be inversely proportional to the total time the ship is in port; a stacker-reclaimer conflict penalty mechanism is called to optimize the fitness function; the stacker-reclaimer conflict penalty mechanism is set to: the greater the number of stacker-reclaimer conflicts, the greater the corresponding penalty value; the stacker-reclaimer conflict is set to: if at a certain time point, multiple tasks use the same stacker-reclaimer at the same time; the stacker-reclaimer conflict penalty value P in the reclaimer conflict penalty mechanism is conflict for: In formula (14), n is the number of conflicting stackers; C j is the number of conflicts of stacker-reclaimer j; λ is the penalty coefficient of conflict, which is a positive integer, so that the more conflicts there are, the greater the penalty; The fitness function is: In formula (15), P conflict is the penalty value for stacker-reclaimer conflict, is the actual arrival time of vessel v at the port, The time when the operation of vessel v is completed and leaves the port.

8. A coal port loading and unloading production intelligent scheduling implementation system, characterized by: Applicable to a coal port including at least two ship loaders, at least two ship unloaders, at least two stackers, at least two loading berths, and at least two unloading berths, the implementation system includes a loading and unloading mixed scheduling optimization model and an improved genetic algorithm: The loading and unloading hybrid scheduling optimization model integrates a model objective function, a combination of model constraints for limiting the model objective function, port yard inventory information, loading and unloading process information, loading and unloading equipment status, and information on ships scheduled to arrive at the port; the model objective function is used to reflect the shortest stay time of all ships at the coal port within the scheduling period, and the stay time at the coal port includes the waiting time for berthing after the ship arrives at the anchorage, the berth operation preparation time, the loading and unloading operation time, and the auxiliary operation time; The improved genetic algorithm is a genetic algorithm improved by a multi-dimensional vector encoding mechanism, population initialization based on process constraint rules, and a fitness function that calls a conflict penalty mechanism. It is used to solve the loading and unloading mixed scheduling optimization model to obtain the optimal scheduling plan; the optimal scheduling plan includes the optimal ship schedule, stacking allocation rules, loading and unloading process arrangement rules, and berth allocation rules; the ship schedule includes the docking time, the start time of operation, and the departure time; the stacking allocation rules include the allocation of stacking spaces in the yard corresponding to the stacker-reclaimer for each ship; the loading and unloading process arrangement rules include the loading and unloading process and equipment usage of each ship; and the berth allocation rules include the allocation of berths for ship operations; The model constraint combination includes: A first model constraint condition, used to constrain the model objective function so as to set a plurality of first restriction rules for determining unique operating conditions for the ship operation process; A second model constraint condition is used to constrain the model objective function so as to set a plurality of second restriction rules for the stacking site resource status parameters; a third model constraint condition, used to constrain the model objective function so as to set a third restriction rule for berth selection of a ship; The fourth model constraint condition is used to constrain the model objective function so as to set at least two fourth restriction rules for the coal port loading and unloading process flow.

9. An electronic device, characterized in that: The electronic device comprises a memory, a processor, and a computer program stored in the memory and executable by the processor, wherein when the processor executes the computer program, the steps of the method for realizing intelligent scheduling of coal port loading and unloading production are implemented as described in any one of claims 1 to 7; the electronic device is used to control the loading and unloading production scheduling of a coal port.

10. A computer-readable storage medium storing a computer program, wherein: When the computer program is executed by a processor, the steps of the method for realizing intelligent scheduling of coal port loading and unloading production as described in any one of claims 1 to 7 are implemented.

Images