A method and apparatus for hoisting and scheduling a data center module

Abstract

The application provides a data center module hoisting scheduling method and device, and relates to the technical field of computers. The method comprises the following steps: receiving module information of a data center module to be installed and scheduling constraint conditions; determining logical dependency relationships between the modules according to the scheduling constraint conditions, wherein the logical dependency relationships comprise at least one of power supply dependency, refrigeration dependency, space dependency and process dependency; generating an initial scheduling scheme set according to the logical dependency relationships; taking the initial scheduling scheme set as an initial population, performing global optimization through a multi-objective genetic algorithm, and outputting an installation scheduling scheme. The application can significantly improve hoisting scheduling efficiency, tower crane utilization rate and anti-disturbance capability, and realize a leap from artificial experience to intelligent optimization.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to a method and apparatus for scheduling the hoisting of data center modules. Background Technology

[0002] Currently, modular construction of data centers involves prefabricating functional units such as power supply, cooling, and IT cabinets into standardized modules in the factory, which are then transported to the site for hoisting and installation. Among existing scheduling methods, solutions based on human experience struggle to manage complex dependencies and are inefficient; solutions based on standard genetic algorithms have poor initial solution quality and slow convergence; and solutions based on global rescheduling are time-consuming to reschedule.

[0003] Therefore, there is an urgent need for an intelligent hoisting scheduling method to solve the problems of existing technologies, such as the inability to handle logical dependency constraints, low quality of initial solutions, neglect of channel blockages, and low efficiency of disturbance response. Summary of the Invention

[0004] This invention provides a data center module hoisting scheduling method and apparatus. This method can solve the problems of logical dependency constraints that cannot be handled, low quality of initial solutions, ignored channel blockages, and low disturbance response efficiency in the prior art, and significantly improve hoisting efficiency, shorten the total construction period, and enhance the robustness of the scheduling scheme.

[0005] Firstly, a data center module hoisting and scheduling method is provided, including: Receive module information of the data center module to be installed, the data center module includes at least one of the power supply module, cooling module and IT module; Receive scheduling constraints, which include module parameters, equipment parameters, site parameters, and time constraints of the data center module; Based on scheduling constraints, determine the logical dependencies between data center modules; Based on logical dependencies, an initial scheduling scheme set is generated, which includes multiple scheduling sequences for installing the data center module; Using the initial set of scheduling schemes as the initial population, a multi-objective genetic algorithm is used to globally optimize the scheduling sequences in the initial population, and output the installation scheduling scheme of the data center module.

[0006] In this way, by establishing logical dependencies between modules and generating an initial scheduling scheme based on these dependencies, and then using a genetic algorithm for global optimization, the problems of ignoring logical dependencies and low quality of initial schemes in existing technologies are solved, significantly improving the feasibility and efficiency of the scheduling scheme.

[0007] In some possible implementations, an initial set of scheduling schemes is generated based on logical dependencies, including: A directed acyclic graph is constructed based on logical dependencies, where nodes in the directed acyclic graph represent data center modules and directed edges represent dependencies between modules. An initial scheduling scheme set is generated using heuristic rules, including logical dependency priority rules. The logical dependency priority rules calculate the priority of each module based on the module's hierarchy and the number of subsequent modules in the directed acyclic graph, and generate a scheduling sequence in descending order of priority.

[0008] In this way, by explicitly modeling dependencies using a directed acyclic graph and generating an initial scheduling sequence using logical dependency priority rules, the initial solution naturally satisfies logical constraints, providing a high-quality starting point for the genetic algorithm and significantly reducing the number of subsequent iterations.

[0009] In some possible implementations, the logical dependencies between data center modules are determined based on scheduling constraints, including: Based on the module type in the module parameters, the power supply module is determined to be installed before the IT modules that depend on it for power supply, thus generating a power supply dependency. Based on the module type in the module parameters, the cooling module is determined to be installed before the IT modules that depend on it for cooling, thus generating a cooling dependency. Based on the relative positions of the edge and center of the hoisting area in the site parameters and the tower crane working radius in the equipment parameters, it is determined that the modules located at the edge of the hoisting area will be hoisted before the modules located in the center area, thus generating spatial dependencies; Based on the time constraints for hoisting a single module and the time required for post-installation debugging, it is determined that the module whose hoisting is completed will start before the subsequent debugging work, thus generating process dependencies.

[0010] In this way, by clearly defining four types of business logic dependencies, the strong constraints in the installation of data center modules are fully covered, ensuring that the scheduling scheme will not be hindered by dependency conflicts in actual execution.

[0011] In some possible implementations, an initial set of scheduling schemes is used as the initial population. A multi-objective genetic algorithm is used to globally optimize the scheduling sequences in the initial population to output the installation scheduling scheme for the data center module, including: Construct a fitness function that includes a total project duration index, a tower crane idle rate index, and a channel accessibility penalty factor. The channel accessibility penalty factor is used to penalize scheduling sequences that cause channel blockage. Calculate the fitness value for each scheduling sequence in the initial population; The tournament selection method is used to select individuals with high fitness values ​​from the current population to enter the next generation of the population. Selected individuals are crossed using a sequential crossover operation to generate a child scheduling sequence, where the sequential crossover operation is used to maintain logical dependencies. The offspring scheduling sequence is mutated by a swap mutation operation, randomly swapping the positions of two non-dependent modules, where a non-dependent module is a module that does not have a direct or indirect dependency relationship in the logical dependency relationship. Repeat the selection, crossover, and mutation operations until the termination condition is met, and output the scheduling sequence with the highest fitness value as the installation scheduling scheme for the data center module.

[0012] In this way, by introducing a channel accessibility penalty factor, the impact of channel blockage during hoisting on subsequent operations is effectively avoided; at the same time, sequential crossover and exchange mutation maintain the dependency relationship, ensuring the effectiveness of the solution during the evolution process, thereby significantly improving the overall quality of the scheduling scheme.

[0013] Among some possible implementations, the method also includes: Execute the installation scheduling plan and monitor the execution status in real time. The execution status includes at least one of the following: tower crane working status data, module arrival time data, and environmental parameter data. When a disturbance event is detected, dynamic rescheduling is triggered.

[0014] In this way, by monitoring the execution status in real time and triggering dynamic rescheduling, the system is given the ability to cope with on-site disturbances, thus enhancing the robustness of the scheduling scheme.

[0015] In some possible implementations, dynamic rescheduling is triggered when a disturbance event is detected, including: When the delay time of the actual arrival time of a module relative to the planned time, determined based on the module arrival time data, is less than the first delay threshold, or when the timeout duration of the hoisting operation of a single module is less than the first timeout threshold, a local adjustment is triggered. A local adjustment means only modifying the adjacent positions of the modules affected by the disturbance in the scheduling sequence. When the delay duration is greater than the first delay threshold but less than the second delay threshold, or when a temporary fault of a single tower crane is detected based on the tower crane's working status data, regional rescheduling is triggered. Regional rescheduling refers to re-optimizing the scheduling of all modules that have not yet been hoisted within the target area affected by the disturbance; wherein, the first delay threshold is less than the second delay threshold. When the delay duration exceeds the second delay threshold, or when multiple tower cranes are detected to be malfunctioning simultaneously based on the tower crane working status data, or when the environmental parameter data does not meet the preset environmental data, a global rescheduling is triggered. Global rescheduling refers to the global optimization of all modules that have not yet been hoisted.

[0016] In this way, by classifying disturbances and using differentiated responses, we can avoid triggering global recalculation by minor disturbances, significantly reduce the time spent on rescheduling, and at the same time ensure the quality of scheme reconstruction under severe disturbances.

[0017] In some possible implementations, triggering a regional rescheduling or a global rescheduling includes: Lock the scheduling sequence of completed modules and extract the set of unexecuted modules; Update the scheduling constraints of the set of unexecuted modules; For the set of unexecuted modules, an initial scheduling scheme set is generated based on logical dependencies. Then, using this initial scheduling scheme set as the initial population, a multi-objective genetic algorithm is used for global optimization to generate a new scheduling scheme. The completed module scheduling sequence is merged with the new scheduling scheme to output a complete installation scheduling scheme.

[0018] In this way, by using an incremental rescheduling strategy, only the unexecuted parts are re-optimized, avoiding the repeated calculation of the completed parts, which significantly improves the rescheduling efficiency and reduces the rescheduling time from tens of minutes in traditional methods to several minutes.

[0019] Secondly, a data center module hoisting and scheduling device is provided, comprising: The first processing module is used to receive module information of the data center module to be installed. The data center module includes at least one of a power supply module, a cooling module, and an information technology equipment module. The second processing module is used to receive scheduling constraints, which include module parameters, equipment parameters, site parameters and time constraints of the data center modules, and determine the logical dependencies between data center modules based on the scheduling constraints. The third processing module is used to generate an initial scheduling scheme set based on logical dependencies, wherein the initial scheduling scheme set includes multiple scheduling sequences for installing the data center module; The fourth processing module is used to take the initial scheduling scheme set as the initial population, perform global optimization on the scheduling sequence in the initial population through a multi-objective genetic algorithm, and output the installation scheduling scheme of the data center module.

[0020] In some possible implementations, the first processing module is specifically used to: obtain module information from the project management system, including the number, type, size, weight, and estimated arrival time of the power supply module, cooling module, and information technology equipment module; obtain equipment parameters from the tower crane control system, including the number of tower cranes, location coordinates, lifting capacity, working radius, and moving speed; obtain site parameters from the site digital map, including the lifting area boundary, main passage location and size, and secondary passage layout; and obtain time constraints from the schedule planning document, including the total project duration limit and key node time requirements.

[0021] In some possible implementations, the second processing module is specifically used to: determine, based on the module type in the module parameters, that the power supply module is hoisted before the IT modules that depend on it for power supply, thus generating a power supply dependency; determine, based on the module type, that the cooling module is hoisted before the IT modules that depend on it for cooling, thus generating a cooling dependency; determine, based on the relative positions of the edge and center of the hoisting area in the site parameters and the tower crane working radius in the equipment parameters, that the modules located at the edge are hoisted before the modules located at the center, thus generating a spatial dependency; and determine, based on the hoisting operation time of a single module and the time required for commissioning after module installation in the time constraints, that the modules that have been hoisted start before the subsequent commissioning operations, thus generating a process dependency.

[0022] In some possible implementations, the third processing module is specifically used to: establish a directed acyclic graph based on logical dependencies, where nodes in the directed acyclic graph represent data center modules and directed edges represent dependencies between modules; generate an initial set of scheduling schemes using heuristic rules, including logical dependency priority rules, which calculate the priority of each module based on the module hierarchy and the number of subsequent modules in the directed acyclic graph, and generate a scheduling sequence in descending order of priority.

[0023] In some possible implementations, the fourth processing module is specifically used for: constructing a fitness function, which includes a total project duration index, a tower crane idle rate index, and a channel accessibility penalty factor, where the channel accessibility penalty factor is used to penalize scheduling sequences that cause channel blockage; calculating the fitness value for each scheduling sequence in the initial population; selecting individuals with high fitness values ​​from the current population to enter the next generation population using a tournament selection method; performing a sequential crossover operation on the selected individuals to generate offspring scheduling sequences, where the sequential crossover operation is used to maintain logical dependencies; performing a swap mutation operation on the offspring scheduling sequences, randomly swapping the positions of two non-dependent modules, where non-dependent modules are modules that do not have direct or indirect dependencies in the logical dependencies; and repeating the selection, crossover, and mutation operations until the termination condition is met, outputting the scheduling sequence with the highest fitness value as the installation scheduling scheme for the data center module.

[0024] In some possible implementations, the device further includes: a monitoring module, configured to trigger a local adjustment when the delay time of the actual arrival time of the module relative to the planned time, determined based on the module arrival time data, is less than a first delay threshold, or the hoisting operation timeout of a single module is less than a first timeout threshold. The local adjustment refers to modifying only the adjacent positions of the disturbed modules in the scheduling sequence. When the delay duration is greater than the first delay threshold and less than the second delay threshold, or when a temporary fault of a single tower crane is detected based on the tower crane's working status data, a regional rescheduling is triggered. The regional rescheduling refers to re-optimizing the scheduling of all modules that have not yet been hoisted within the target area affected by the disturbance; wherein, the first delay threshold is less than the second delay threshold. When the delay duration exceeds the second delay threshold, or when multiple tower cranes are detected to be malfunctioning simultaneously based on the tower crane working status data, or when the environmental parameter data does not meet the preset environmental data, a global rescheduling is triggered. The global rescheduling refers to the global optimization of all modules that have not yet been hoisted.

[0025] Thirdly, an electronic device is provided, comprising: one or more processors; one or more memories; and one or more programs, wherein the one or more programs are stored in the one or more memories, and the one or more programs include instructions that, when executed by the one or more processors, cause the configured device to perform the method as described in the first aspect.

[0026] Fourthly, a computer storage medium is provided, the computer storage medium storing instructions that, when executed by a computer, cause the computer to perform the method described in the first aspect or the second aspect.

[0027] Fifthly, a computer program product is provided, the computer program product storing instructions that, when executed by a computer, cause the computer to perform the method described in the first aspect or the second aspect.

[0028] The beneficial effects of the second to fifth aspects can be referred to the introduction of the beneficial effects of the first aspect above, and will not be repeated here. Attached Figure Description

[0029] Figure 1 This is an overall architecture block diagram of a data center module hoisting and scheduling method according to an embodiment of the present invention; Figure 2 This is a flowchart illustrating a data center module hoisting and scheduling method according to an embodiment of the present invention; Figure 3 A flowchart for generating an initial solution provided in an embodiment of the present invention; Figure 4 This is a flowchart of the incremental rescheduling process provided in an embodiment of the present invention; Figure 5 A flowchart of a data center module hoisting and scheduling method provided in an embodiment of the present invention; Figure 6This is a schematic diagram of a data center module hoisting and scheduling device that can be used to implement the method of the present invention; Figure 7 This is a schematic diagram of an electronic device provided by the present invention. Detailed Implementation

[0030] The solutions provided by the embodiments of the present invention will now be described with reference to the accompanying drawings. In the embodiments of the present invention, "multiple" refers to two or more objects, and "various kinds" refers to two or more types. Terms such as "first," "second," etc., are only used to distinguish similar objects and are not necessarily used to describe a specific order or number of objects.

[0031] Currently, modular construction of data centers involves prefabricating functional units such as power supply, cooling, and IT cabinets into standardized modules in the factory, which are then transported to the site for hoisting and installation. In typical projects with hundreds of modules, multiple tower cranes are involved in coordinated operations, and scheduling efficiency directly impacts the project schedule and cost. However, existing mainstream scheduling methods rely on manual experience or general optimization algorithms, which have the following significant limitations.

[0032] Scheduling schemes based on human experience rely on the experience of on-site management personnel to prioritize modules and allocate tower cranes. However, manual scheduling struggles to manage the complex dependencies between dozens or even hundreds of modules, resulting in generally low tower crane utilization and often exceeding the planned timeframe. Furthermore, manual adjustments are slow to respond to sudden disruptions such as module delays and equipment failures, easily causing process blockages and downtime.

[0033] Optimization schemes based on standard genetic algorithms, as shown in comparison documents 1-3, model hoisting scheduling as a combinatorial optimization problem and use genetic algorithms to output module sequences and time plans. However, these general-purpose algorithms do not deeply consider the unique business logic dependencies between data center modules. For example, power supply modules must be in place before IT modules can be powered on and debugged, and cooling modules must be precisely aligned with row-level air conditioners. Such existing technologies typically use fixed priorities or completely randomized initial schemes, which cannot effectively handle the above strong logical constraints. This leads to frequent obstacles in the generated scheduling sequence during actual execution, resulting in a low optimal solution finding rate.

[0034] The disturbance response scheme based on global rescheduling adopts a rescheduling strategy that recalculates all unexecuted modules when disturbances such as equipment failure and module delays are detected. Because it cannot distinguish the disturbance level (minor, moderate, or major), even a slight delay triggers a full calculation, resulting in excessively long rescheduling times and severely hindering on-site progress. Furthermore, the existing fitness function primarily focuses on minimizing the total project duration and maximizing resource utilization, neglecting the dynamic impact of accessibility during hoisting on subsequent operations. Actual measurements show that access blockages significantly increase the waiting time for subsequent modules.

[0035] In summary, existing technologies lack the ability to model logical dependencies between modules, the low quality of initial solutions leads to slow algorithm convergence, the fitness function does not consider dynamic path blocking, and the disturbance response lacks a hierarchical mechanism. Specifically: manual scheduling is inefficient and highly volatile; standard genetic algorithms cannot handle strong logical dependencies such as power supply and cooling; global rescheduling schemes are overly burdensome and time-consuming for minor disturbance responses; and none of them have constructed an intelligent hoisting scheduling closed-loop system that integrates dependency constraint modeling, heuristic initial solution generation, channel blocking penalties, and hierarchical rescheduling, resulting in project delays, idle tower cranes, and high costs.

[0036] To address the technical problems of existing technologies, such as the inability to handle logical dependency constraints, low quality of initial solutions, neglect of path congestion, and low efficiency in disturbance response, this invention proposes a data center module hoisting and scheduling method and system for modular data center construction scenarios. By establishing a directed acyclic graph of module dependencies, business logic constraints such as power supply and cooling dependencies are incorporated into the scheduling algorithm. Three heuristic rules—logical dependency priority, dynamic path congestion evaluation, and value-time weighting—are used to generate high-quality initial solutions, significantly reducing the number of iterations required for convergence. A channel availability penalty factor is introduced into the fitness function, effectively shortening the total project duration. Local, regional, or global rescheduling is performed based on the disturbance level, significantly reducing rescheduling time.

[0037] This method achieves a leap from manual experience or general algorithms to intelligent scheduling of constraint modeling, heuristic initial solutions, adaptive optimization, and hierarchical disturbance response, significantly improving the efficiency, robustness, and automation level of data center module hoisting.

[0038] Figure 1 This is an overall architecture block diagram of a data center module hoisting and scheduling method shown in an embodiment of the present invention. Figure 1 As shown, the overall architecture consists of four parts: the input layer, the core processing layer, the output layer, and the monitoring layer.

[0039] The constraints required for input layer data acquisition and scheduling include module parameters, equipment parameters, site parameters, and time constraints. Module parameters include module number, type, size, weight, and arrival time. Power supply modules also include rated power and interface specifications; cooling modules include cooling capacity and interface size; and IT modules include rack height and load-bearing requirements. Equipment parameters include the number of tower cranes, their location coordinates, lifting capacity, working radius, and moving speed. Site parameters include the lifting area boundaries, main channel location and dimensions, secondary channel layout, and occupancy information of already positioned modules. Time constraints include the total project duration limit, critical node time requirements, and daily work time windows.

[0040] The core processing layer communicates with the input layer and is the computational core of the system. It includes a dependency modeling module, a heuristic rule engine, a genetic algorithm optimization module, and a fitness function calculation module.

[0041] The dependency modeling module constructs a directed acyclic graph (DAG) based on the logical dependencies between modules. Logical dependencies include power supply dependencies, cooling dependencies, spatial dependencies, and process dependencies. Power supply dependencies mean the power supply module must be in place before the IT module can be powered on and debugged; cooling dependencies mean the cooling module must be precisely aligned with the row-level air conditioner; spatial dependencies mean peripheral modules must be in place before internal modules to ensure unobstructed passageways; and process dependencies mean debugging can only begin after installation is complete. This module outputs a dependency graph G=(V,E), where V is the set of modules and E is the set of dependency edges.

[0042] The heuristic rule engine uses three heuristic rules to generate a high-quality set of initial scheduling schemes, instead of the random generation used in traditional genetic algorithms. The three rules are: a logical dependency priority rule, which calculates priority based on the module's hierarchy in the dependency graph and the number of subsequent modules, ensuring dependent modules are prioritized; a dynamic path congestion evaluation rule, which calculates the congestion coefficient of each module on the main channel, prioritizing modules with lower congestion coefficients to reduce waiting time for subsequent modules; and a value-time weighted rule, which considers both the module's value coefficient and the ratio of debugging cycle to scheduling time, prioritizing high-value modules with long debugging cycles. The heuristic rule engine outputs N initial scheduling sequences as the initial population for the genetic algorithm.

[0043] The genetic algorithm optimization module starts with the initial population and performs global optimization using a multi-objective genetic algorithm. This module employs genetic operations such as tournament selection, sequential crossover, and crossover mutation. Sequential crossover preserves dependencies, while crossover mutation randomly swaps the positions of non-dependent modules.

[0044] The fitness function calculation module is designed specifically for fitness functions used in data center module hoisting scenarios.

[0045] The output layer communicates with the core processing layer to output optimal or suboptimal scheduling schemes. Output includes scheduling sequences, time plans, resource allocation, and path planning. The output scheme is presented in the form of a visual Gantt chart, table, or structured instructions, and can be pushed to the field management terminal and tower crane control system.

[0046] The monitoring layer communicates with both the output and input layers, responsible for executing the scheduling plan and monitoring the execution status in real time. When a disturbance event is detected, dynamic rescheduling is triggered. The monitoring layer includes a disturbance detection module and a rescheduling trigger module. The disturbance detection module collects execution data, calculates indicators such as the deviation between actual and planned progress, actual tower crane utilization, and module arrival delays, and classifies disturbance events into Level 1, Level 2, and Level 3 disturbances. The rescheduling trigger module executes differentiated responses based on the disturbance level: Level 1 disturbances perform local adjustments, only adjusting adjacent sequences of affected modules; Level 2 disturbances perform regional rescheduling, adjusting unexecuted modules within the affected region; Level 3 disturbances perform global rescheduling, re-optimizing all unexecuted modules. Rescheduling employs an incremental strategy, locking the scheduling sequences of completed modules and only re-optimizing the set of unexecuted modules, merging them to generate a complete new plan. This mechanism reduces rescheduling time from tens of minutes in traditional methods to several minutes.

[0047] Through the above overall framework, an intelligent scheduling leap has been achieved, moving from human experience or general algorithms to constraint modeling, heuristic initialization, adaptive optimization, and hierarchical disturbance response.

[0048] Figure 2 This is a flowchart illustrating a data center module hoisting and scheduling method according to an embodiment of the present invention. Figure 2 As shown, the method includes inputting constraints and establishing a dependency graph, generating an initial scheduling scheme using heuristic rules, feeding it into a genetic algorithm for global optimization, outputting the scheme if the termination condition is met, otherwise continuing optimization; real-time monitoring during scheme execution, and triggering dynamic rescheduling if a disturbance event is detected, returning to recalculation to form a closed loop.

[0049] The following is a detailed explanation of each step.

[0050] In step S1, the scheduling constraints are input.

[0051] In this embodiment of the invention, the input parameters include module parameters, equipment parameters, site parameters, and time constraints.

[0052] The module parameters include number, type, size, weight, and arrival time; the equipment parameters include the number, location, and lifting capacity of tower cranes; the site parameters include the hoisting area and access location; and the time constraints include the total construction period and key milestones.

[0053] The set of constraints is represented as C = {C1, C2, C3, ..., Cn}, where C1 is a module. i The arrival time is no greater than its scheduling start time; C2 is a module. i The weight is no greater than that of the tower crane. j Maximum lifting capacity; C3 is a module iThe hoisting time shall not exceed the available time window; C4 indicates that if dependencies exist, the dependent modules... k Must precede module i The hoisting is complete.

[0054] In step S2, a module dependency graph is established.

[0055] In this embodiment of the invention, a directed acyclic graph is constructed. , For a collection of modules, It is a set of dependencies.

[0056] like , belong ,but Must precede Hoisting. Dependency types include power supply dependency (power supply module precedes IT module), cooling dependency (cooling module precedes IT module), space dependency (edge ​​module precedes center module), and process dependency (hoisting completion precedes commissioning).

[0057] Specifically, in the modular construction of data centers, IT modules refer to standardized functional units that house information technology equipment, typically including server racks, network equipment racks, and storage equipment racks. IT modules are the core computing resource carriers of a data center, responsible for data processing, storage, and transmission.

[0058] In hoisting scheduling scenarios, IT modules have the following characteristics: Attribute parameters include module number, type (IT), size, weight, power consumption, and heat dissipation requirements. For example, a standard IT cabinet is typically 600 mm wide, has a height in U units, and weighs between 500 kg and 1500 kg.

[0059] Logical dependencies: The IT module depends on the power supply module and the cooling module. The power supply module must be in place before the IT module; otherwise, the IT module cannot be powered on and debugged. The cooling module must be in place before the IT module to ensure that the heat generated by the IT equipment during operation can be cooled in a timely manner.

[0060] Scheduling priority: In the value-time weighted rule, the value coefficient of IT modules is usually 1.0. They have a shorter debugging cycle and relatively less hoisting time, so their priority is usually lower than that of power supply modules and cooling modules.

[0061] In step S3, an initial scheduling scheme is generated using heuristic rules.

[0062] In this embodiment of the invention, three heuristic rules are used to generate a high-quality initial solution, rather than generating it randomly.

[0063] Rule 1: Logical dependencies take precedence. If a dependency exists... ,but The priority is higher than The priority calculation formula is as follows: , Where Base_Priority = 100 - the level of module i in the dependency graph × 10, and Dependency_Bonus = the number of subsequent modules of dependent module i × 5.

[0064] Rule 2: Dynamic path blocking assessment rule.

[0065] Path blocking coefficient ,in Let be the distance between module i and main channel k. The reference distance for the main passage k (usually half the width of the site). The weight of module i for channel k (1.5 for large modules, 1.0 for small modules). The smaller the value, the higher the priority.

[0066] Rule 3: Value-Time Weighted Rule.

[0067] Module Priority ,in, The module value coefficients are: power supply module 1.5, cooling module 1.3, IT module 1.0. The module debugging cycle (days). The time (in hours) for module hoisting. The larger the value, the higher the priority.

[0068] Figure 3 A flowchart illustrating the heuristic initial solution generation process is shown, such as... Figure 3 As shown, the process includes the following steps: Step 1, calculate the dependency priority of each module; Step 2, calculate the path blocking coefficient of each module; Step 3, calculate the value time weight of each module; Step 4, generate an initial ranking by combining the scores of the above three items; Step 5, generate N initial scheduling sequences based on the initial ranking, which serve as the initial population of the genetic algorithm.

[0069] In step S4, the multi-objective genetic algorithm performs global optimization.

[0070] Specifically, the algorithm parameters are set as follows: population size 100 to 200, crossover probability 0.8 to 0.9, mutation probability 0.05 to 0.1, and maximum number of iterations 300 to 500.

[0071] The fitness function is designed as follows: ,in For the total construction period, The total idle rate of tower cranes is calculated by dividing the sum of the idle times of all tower cranes by the total available time. This is a penalty factor for channel unobstructedness.

[0072] The dynamic adjustment strategy for weighting coefficients is as follows: in the early stage of the project, α=0.3, β=0.3, γ=0.4, prioritizing ensuring smooth access; in the middle stage of the project, α=0.4, β=0.4, γ=0.2, balancing the construction period and resources; in the later stage of the project, α=0.5, β=0.3, γ=0.2, prioritizing accelerating the completion of tasks.

[0073] Channel Flow Penalty Factor ,in For channel k The degree of congestion (0 to 1). The importance coefficients for the channels are (1.5 for the primary channel and 1.0 for the secondary channel).

[0074] In genetic operations, the selection operation uses tournament selection, randomly selecting 5 individuals each time, and choosing the one with the highest fitness to enter the next generation; the crossover operation uses sequential crossover to maintain the dependency relationship; and the mutation operation uses exchange mutation, randomly swapping the positions of two independent modules.

[0075] In step S5, the termination condition is determined.

[0076] In this embodiment of the invention, the process terminates when any of the following conditions are met: the number of iterations reaches a preset maximum value, such as 500 times; the optimal fitness value does not improve significantly for 50 consecutive generations, and the improvement is less than one percent; or a solution that satisfies all hard constraints and has a fitness greater than the threshold is found.

[0077] In step S6, the optimal or suboptimal scheduling scheme is output.

[0078] In this embodiment of the invention, the output includes scheduling sequence, time plan, resource allocation and path planning.

[0079] The output format example is as follows: Scheduling scheme ID: SCH-2026-001; the module scheduling sequence includes the sequence number, module number, module type, tower crane number, start time and end time; expected indicators include total construction period of 56 days and average tower crane utilization rate of 87%.

[0080] In step S7, scheduling is performed and real-time monitoring is conducted.

[0081] In this embodiment of the invention, the monitoring data acquisition methods include: module location data collected every 10 seconds via UWB positioning tags; tower crane status data collected every 5 seconds via the tower crane black box; environmental parameters data collected every minute via on-site weather stations; and manual progress data entered in real time via mobile terminals. Monitoring indicators include the deviation between actual and planned progress, the comparison between actual and planned tower crane utilization rates, and the comparison between module arrival time and estimated arrival time.

[0082] In step S8, disturbance events are detected.

[0083] The disturbance events are classified as follows: Level 1 disturbances are module delays of less than 1 hour or single module hoisting timeouts of less than 30 minutes, triggering local adjustments; Level 2 disturbances are module delays of 1 to 4 hours or temporary failures of a single tower crane, triggering regional rescheduling; Level 3 disturbances are module delays of more than 4 hours, multiple tower crane failures, or severe weather, triggering global rescheduling.

[0084] Specifically, when the delay time of the actual arrival time of a module relative to the planned time, determined based on the module arrival time data, is less than the first delay threshold, or when the timeout duration of the hoisting operation of a single module is less than the first timeout threshold, a local adjustment is triggered. Local adjustment refers to modifying only the adjacent positions of the modules affected by the disturbance in the scheduling sequence. When the delay duration is greater than the first delay threshold and less than the second delay threshold, or when a temporary fault of a single tower crane is detected based on the tower crane's working status data, regional rescheduling is triggered. Regional rescheduling refers to re-optimizing the scheduling of all modules that have not yet been hoisted within the target area affected by the disturbance; wherein, the first delay threshold is less than the second delay threshold. When the delay duration exceeds the second delay threshold, or multiple tower cranes are detected to be malfunctioning simultaneously based on tower crane operating status data, or environmental parameter data does not meet preset environmental data, a global rescheduling is triggered. Global rescheduling refers to a global re-optimization of all modules that have not yet been hoisted. In step S9, if no disturbance event occurs, the scheduling scheme is output.

[0085] In step S10, dynamic rescheduling is triggered.

[0086] This invention employs incremental rescheduling, optimizing only the M modules that have not been executed, rather than recalculating all N modules.

[0087] Figure 4 This is a flowchart illustrating the incremental rescheduling process according to an embodiment of the present invention. Figure 4As shown, the process includes the following steps: after detecting a disturbance, first check the completed modules and lock the scheduling sequence of the completed modules; extract the set of unexecuted modules, update the constraints and currently available resources; execute the optimization algorithm on the unexecuted modules; finally merge the scheduling sequence of the completed modules with the newly optimized scheme and output the complete new scheduling scheme.

[0088] In practical applications, for a data center project comprising 100 modules (20 power supply modules, 30 cooling modules, and 50 IT modules), 4 tower cranes were configured, with a planned construction period of 60 days. The comparative test results are as follows: Manually scheduled tower cranes achieved a utilization rate of 62%, a total construction period of 78 days, a disturbance impact period of 12 days, and a rescheduling response time of 120 minutes; the pure genetic algorithm achieved a tower crane utilization rate of 75%, a total construction period of 65 days, a disturbance impact period of 8 days, 650 convergence iterations, a rescheduling response time of 45 minutes, and an optimal solution finding rate of 65%; the hybrid algorithm of this invention increased tower crane utilization to 87%, shortened the total construction period to 56 days, reduced the disturbance impact period to 3 days, reduced the number of convergence iterations to 280, shortened the rescheduling response time to 8 minutes, and achieved an optimal solution finding rate as high as 92%.

[0089] Furthermore, in another practical application, for a large data center project comprising 200 modules, 8 tower cranes were configured, with a planned construction period of 120 days. Test results show that compared to the traditional solution, the total construction period of the present invention was shortened from 156 days to 128 days, an improvement of 18%; tower crane idle time was reduced from 2400 hours to 1200 hours, a reduction of 50%; channel congestion events were reduced from 28 times to 6 times, a reduction of 79%; and the average rescheduling time was reduced from 55 minutes to 9 minutes, a reduction of 84%.

[0090] The above embodiments fully demonstrate the significant advantages of the present invention in improving tower crane utilization, shortening the overall construction period, reducing disturbance impact, accelerating convergence speed, reducing rescheduling time, and improving the optimal solution finding rate, especially in large-scale projects where the effect is more prominent.

[0091] In summary, this invention effectively solves the problem of existing technologies being unable to handle logical dependency constraints by establishing a module dependency graph and incorporating business logic constraints such as power supply and cooling dependencies into the scheduling algorithm. It significantly improves the algorithm's convergence speed by using heuristic rules to generate high-quality initial solutions. Furthermore, it effectively shortens the overall project duration and optimizes path congestion during hoisting by introducing a channel accessibility penalty factor into the fitness function. Finally, it significantly reduces rescheduling time by performing local, regional, or global rescheduling based on disturbance levels, achieving hierarchical disturbance response and significantly improving the robustness and execution efficiency of the scheduling scheme.

[0092] Figure 5A flowchart of a data center module hoisting and scheduling method provided in an embodiment of the present invention is shown below. Figure 5 As shown, it includes: 510: Receive module information for the data center module to be installed. The data center module includes at least one of a power supply module, a cooling module, and an IT module.

[0093] 520: Receive scheduling constraints, which include module parameters, equipment parameters, site parameters, and time constraints of the data center modules; determine the logical dependencies between data center modules based on the scheduling constraints.

[0094] 530: Generate an initial scheduling scheme set based on logical dependencies, which includes multiple scheduling sequences for installing the data center module.

[0095] 540: Using the initial set of scheduling schemes as the initial population, a multi-objective genetic algorithm is used to globally optimize the scheduling sequences in the initial population, and output the installation scheduling scheme of the data center module.

[0096] By establishing logical dependencies between modules and generating an initial scheduling scheme based on these dependencies, and then using a genetic algorithm for global optimization, the problems of ignoring logical dependencies and low quality of initial schemes in existing technologies are solved, significantly improving the feasibility and efficiency of the scheduling scheme.

[0097] In some embodiments, step 530 includes: A directed acyclic graph is constructed based on logical dependencies, where nodes in the directed acyclic graph represent data center modules, and directed edges represent dependencies between modules.

[0098] An initial scheduling scheme set is generated using heuristic rules, including logical dependency priority rules. The logical dependency priority rules calculate the priority of each module based on the module's hierarchy and the number of subsequent modules in the directed acyclic graph, and generate a scheduling sequence in descending order of priority.

[0099] By explicitly modeling dependencies using a directed acyclic graph and generating an initial scheduling sequence using a logical dependency priority rule, the initial solution naturally satisfies the logical constraints, providing a high-quality starting point for the genetic algorithm and significantly reducing the number of subsequent iterations.

[0100] In some embodiments, step 520, determining the logical dependencies between data center modules based on scheduling constraints, includes: Based on the module type in the module parameters, the power supply module is determined to be installed before the IT modules that depend on it for power supply, thus generating a power supply dependency.

[0101] Based on the module type in the module parameters, the cooling module is installed before the IT modules that depend on it for cooling, thus generating a cooling dependency.

[0102] Based on the relative positions of the edge and center of the hoisting area in the site parameters and the tower crane working radius in the equipment parameters, it is determined that the modules located at the edge of the hoisting area will be hoisted before the modules located in the center area, thus generating spatial dependency.

[0103] Based on the time constraints for hoisting a single module and the time required for post-installation debugging, it is determined that the module whose hoisting is completed will start before the subsequent debugging work, thus generating process dependencies.

[0104] By clearly defining four types of business logic dependencies, the system comprehensively covers the strong constraints in the installation of data center modules, ensuring that the scheduling scheme is not hindered by dependency conflicts during actual execution.

[0105] In some embodiments, step 540 includes: Construct a fitness function that includes a total project duration index, a tower crane idle rate index, and a channel accessibility penalty factor. The channel accessibility penalty factor is used to penalize scheduling sequences that cause channel blockage.

[0106] Calculate the fitness value for each scheduling sequence in the initial population.

[0107] The tournament selection method is used to select individuals with high fitness values ​​from the current population to enter the next generation of the population.

[0108] Selected individuals are cross-crossed using a sequential crossover operation to generate a child scheduling sequence, where the sequential crossover operation is used to maintain logical dependencies.

[0109] The offspring scheduling sequence is mutated by a swap mutation operation, randomly swapping the positions of two non-dependent modules. A non-dependent module is a module that does not have a direct or indirect dependency relationship in the logical dependency relationship.

[0110] Repeat the selection, crossover, and mutation operations until the termination condition is met, and output the scheduling sequence with the highest fitness value as the installation scheduling scheme for the data center module.

[0111] By introducing a channel accessibility penalty factor, the impact of channel blockage during hoisting on subsequent operations is effectively avoided. At the same time, sequential crossover and exchange mutations maintain the dependency relationship, ensuring the effectiveness of the solution during the evolution process, thereby significantly improving the overall quality of the scheduling scheme.

[0112] In some embodiments, it also includes: Execute the installation scheduling plan and monitor the execution status in real time. The execution status includes at least one of the following: tower crane working status data, module arrival time data, and environmental parameter data. When a disturbance event is detected, dynamic rescheduling is triggered.

[0113] By monitoring the execution status in real time and triggering dynamic rescheduling, the system is given the ability to cope with on-site disturbances, thus enhancing the robustness of the scheduling scheme.

[0114] In some embodiments, when a disturbance event is detected, dynamic rescheduling is triggered, including: When the delay between the actual arrival time of a module and the planned time, determined based on the module arrival time data, is less than the first delay threshold, or when the timeout duration of the hoisting operation of a single module is less than the first timeout threshold, a local adjustment is triggered. A local adjustment means modifying only the adjacent positions of the modules affected by the disturbance in the scheduling sequence.

[0115] When the delay duration is greater than the first delay threshold but less than the second delay threshold, or when a temporary fault of a single tower crane is detected based on the tower crane's working status data, regional rescheduling is triggered. Regional rescheduling refers to re-optimizing the scheduling of all modules that have not yet been hoisted within the target area affected by the disturbance; wherein, the first delay threshold is less than the second delay threshold. When the delay duration exceeds the second delay threshold, or when multiple tower cranes are detected to be malfunctioning simultaneously based on the tower crane working status data, or when the environmental parameter data does not meet the preset environmental data, a global rescheduling is triggered. Global rescheduling refers to the global optimization of all modules that have not yet been hoisted.

[0116] In this way, by classifying disturbances and using differentiated responses, we can avoid triggering global recalculation by minor disturbances, significantly reduce the time spent on rescheduling, and at the same time ensure the quality of scheme reconstruction under severe disturbances.

[0117] In some embodiments, triggering a regional rescheduling or a global rescheduling includes: Lock the scheduling sequence of completed modules and extract the set of unexecuted modules.

[0118] Update the scheduling constraints of the set of unexecuted modules.

[0119] For the set of unexecuted modules, an initial scheduling scheme set is generated based on logical dependencies. Then, using this initial scheduling scheme set as the initial population, a multi-objective genetic algorithm is used for global optimization to generate a new scheduling scheme.

[0120] The completed module scheduling sequence is merged with the new scheduling scheme to output a complete installation scheduling scheme.

[0121] By using an incremental rescheduling strategy, only the unexecuted parts are re-optimized, avoiding the repeated calculation of the completed parts, which significantly improves the rescheduling efficiency and reduces the rescheduling time from tens of minutes in traditional methods to several minutes.

[0122] Figure 6 This is a schematic diagram of a data center module hoisting and scheduling device that can be used to implement the method of the present invention, as shown below. Figure 6 As shown, it includes: The first processing module is used to receive module information of the data center module to be installed. The data center module includes at least one of a power supply module, a cooling module, and an information technology equipment module.

[0123] The second processing module is used to receive scheduling constraints, which include module parameters, equipment parameters, site parameters and time constraints of the data center modules, and determine the logical dependencies between data center modules based on the scheduling constraints.

[0124] The third processing module is used to generate an initial scheduling scheme set based on logical dependencies, wherein the initial scheduling scheme set includes multiple scheduling sequences for installing the data center module.

[0125] The fourth processing module is used to take the initial scheduling scheme set as the initial population, perform global optimization on the scheduling sequence in the initial population through a multi-objective genetic algorithm, and output the installation scheduling scheme of the data center module.

[0126] In some embodiments, the first processing module is specifically used to: obtain module information from the project management system, including the number, type, size, weight, and estimated arrival time of the power supply module, cooling module, and information technology equipment module; obtain equipment parameters from the tower crane control system, including the number of tower cranes, location coordinates, lifting capacity, working radius, and moving speed; obtain site parameters from the site digital map, including the lifting area boundary, main passage location and size, and secondary passage layout; and obtain time constraints from the schedule planning document, including the total schedule limit and key node time requirements.

[0127] In some embodiments, the second processing module is specifically used to: determine, based on the module type in the module parameters, that the power supply module is hoisted before the IT modules that depend on it for power supply, thus generating a power supply dependency; determine, based on the module type, that the cooling module is hoisted before the IT modules that depend on it for cooling, thus generating a cooling dependency; determine, based on the relative position of the edge and center of the hoisting area in the site parameters and the tower crane working radius in the equipment parameters, that the modules located at the edge are hoisted before the modules located at the center, thus generating a spatial dependency; and determine, based on the hoisting operation time of a single module and the time required for commissioning after module installation in the time constraints, that the modules that have been hoisted start before the subsequent commissioning operations, thus generating a process dependency.

[0128] In some embodiments, the third processing module is specifically used to: establish a directed acyclic graph based on logical dependencies, wherein the nodes of the directed acyclic graph represent data center modules and the directed edges represent dependencies between modules; generate an initial scheduling scheme set using heuristic rules, including logical dependency priority rules, which calculate the priority of each module based on the module hierarchy and the number of subsequent modules in the directed acyclic graph, and generate a scheduling sequence in descending order of priority.

[0129] In some embodiments, the fourth processing module is specifically used for: constructing a fitness function, which includes a total project duration index, a tower crane idle rate index, and a channel accessibility penalty factor, wherein the channel accessibility penalty factor is used to penalize scheduling sequences that cause channel blockage; calculating the fitness value for each scheduling sequence in the initial population; selecting individuals with high fitness values ​​from the current population to enter the next generation population using a tournament selection method; performing a sequential crossover operation on the selected individuals to generate offspring scheduling sequences, wherein the sequential crossover operation is used to maintain logical dependencies; performing a swap mutation operation on the offspring scheduling sequences to randomly swap the positions of two non-dependent modules, wherein non-dependent modules are modules that do not have direct or indirect dependencies in the logical dependencies; repeating the selection operation, crossover operation, and mutation operation until the termination condition is met, and outputting the scheduling sequence with the highest fitness value as the installation scheduling scheme for the data center module.

[0130] In some embodiments, the apparatus further includes: a monitoring module, configured to trigger a local adjustment when the delay time of the actual arrival time of the module relative to the planned time, determined based on the module arrival time data, is less than a first delay threshold, or the timeout duration of the hoisting operation of a single module is less than a first timeout threshold, wherein the local adjustment refers to modifying only the adjacent positions of the disturbed modules in the scheduling sequence. When the delay duration is greater than the first delay threshold and less than the second delay threshold, or when a temporary fault of a single tower crane is detected based on the tower crane's working status data, a regional rescheduling is triggered. The regional rescheduling refers to re-optimizing the scheduling of all modules that have not yet been hoisted within the target area affected by the disturbance; wherein, the first delay threshold is less than the second delay threshold. When the delay duration exceeds the second delay threshold, or when multiple tower cranes are detected to be malfunctioning simultaneously based on the tower crane operating status data, or when the environmental parameter data does not meet preset environmental data, a global rescheduling is triggered. This global rescheduling refers to a global re-optimization of all modules that have not yet performed hoisting operations. Those skilled in the art should readily recognize that, based on the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein, the present invention can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed in hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present invention.

[0131] It should be noted that, Figure 6 The division of modules / units is illustrative and represents only one logical functional division; in actual implementation, other division methods are possible. For example, two or more functions can be integrated into a single data acquisition module. The integrated modules described above can be implemented either in hardware or as software functional modules.

[0132] An electronic device according to an embodiment of the present invention includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements any of the above-mentioned data center module hoisting and scheduling methods. That is, an electronic device according to an embodiment of the present invention may include, but is not limited to: a processor and a memory; the memory is used to store the computer program; the processor is used to execute the data center module hoisting and scheduling method shown in any embodiment of the present invention by calling the computer program.

[0133] In one alternative embodiment, an electronic device is provided, such as Figure 7 As shown, Figure 7 The illustrated electronic device 700 includes a processor 701 and a memory 703. The processor 701 and the memory 703 are connected, for example, via a bus 702. Optionally, the electronic device 700 may further include a transceiver 704, which can be used for data interaction between the electronic device and other electronic devices, such as sending and / or receiving data. It should be noted that in practical applications, the transceiver 704 is not limited to one type, and the structure of the electronic device 700 does not constitute a limitation on the embodiments of the present invention.

[0134] Among them, electronic devices can also be terminal devices, which can be any device that can install applications, including at least one of smartphones, tablets, laptops, desktop computers, smart speakers, smartwatches, smart TVs, and smart in-vehicle devices.

[0135] It should be noted that, Figure 7 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments of the present invention.

[0136] An embodiment of the present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements any of the above-described data center module hoisting and scheduling methods.

[0137] The above description is merely a preferred embodiment of the present invention and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of disclosure in this invention is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-disclosed concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features with similar functions disclosed in this invention.

[0138] It should be noted that the terms "first," "second," etc., used in the specification and claims of this invention are used to distinguish similar objects and represent a limitation on a specific order or sequence. Where appropriate, the order of use for similar objects can be interchanged so that the embodiments of the invention described herein can be implemented in an order other than that shown or described.

[0139] Those skilled in the art will recognize that this invention can be implemented as a system, method, or computer program product. Therefore, this invention can be specifically implemented in the following forms: it can be entirely hardware, entirely software (including firmware, resident software, microcode, etc.), or a combination of hardware and software, generally referred to herein as a "circuit," "module," or "system." Furthermore, in some embodiments, this invention can also be implemented as a computer program product contained in one or more computer-readable media, which includes computer-readable program code.

[0140] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for scheduling the hoisting of data center modules, characterized in that, include: Receive module information of the data center module to be installed, wherein the data center module includes at least one of a power supply module, a cooling module, and an IT module; The system receives scheduling constraints, which include module parameters, equipment parameters, site parameters, and time constraints of the data center module. Based on the scheduling constraints, the logical dependencies between the data center modules are determined; Based on the logical dependencies, an initial scheduling scheme set is generated, wherein the initial scheduling scheme set includes multiple scheduling sequences for installing the data center module; Using the initial set of scheduling schemes as the initial population, a multi-objective genetic algorithm is used to globally optimize the scheduling sequences in the initial population, and the installation scheduling scheme of the data center module is output.

2. The method according to claim 1, characterized in that, The step of generating an initial scheduling scheme set based on the logical dependencies includes: Based on the logical dependencies, a directed acyclic graph is constructed, wherein the nodes of the directed acyclic graph represent the data center modules, and the directed edges represent the dependencies between modules. The initial scheduling scheme set is generated using heuristic rules, including logical dependency priority rules. The logical dependency priority rules calculate the priority of each module based on the module hierarchy and the number of subsequent modules in the directed acyclic graph, and generate the scheduling sequence in descending order of priority.

3. The method according to claim 1, characterized in that, Determining the logical dependencies between the data center modules based on the scheduling constraints includes: Based on the module type in the module parameters, the power supply module is determined to be installed before the IT modules that depend on it for power supply, thus generating a power supply dependency. Based on the module type in the module parameters, the cooling module is determined to be installed before the IT modules that depend on it for cooling, thus generating a cooling dependency. Based on the relative positions of the edge and center of the hoisting area in the site parameters and the tower crane working radius in the equipment parameters, it is determined that the module located at the edge of the hoisting area will be hoisted before the module located in the center area, thus generating spatial dependency; Based on the time constraints for hoisting a single module and the time required for post-installation debugging of the module, it is determined that the module whose hoisting is completed will begin its subsequent debugging operation first, thus generating process dependencies.

4. The method according to claim 1, characterized in that, The process involves using the initial scheduling scheme set as the initial population, and globally optimizing the scheduling sequences in the initial population using a multi-objective genetic algorithm to output the installation scheduling scheme for the data center module, including: Construct a fitness function, which includes a total project duration index, a tower crane idle rate index, and a channel accessibility penalty factor, wherein the channel accessibility penalty factor is used to penalize scheduling sequences that cause channel blockage; Calculate the fitness value for each scheduling sequence in the initial population; The tournament selection method is used to select individuals with high fitness values ​​from the current population to enter the next generation of the population. Selected individuals are cross-crossed using a sequential cross-crossing operation to generate a child scheduling sequence, wherein the sequential cross-crossing operation is used to maintain the logical dependency relationship. The offspring scheduling sequence is mutated by a swap mutation operation, randomly swapping the positions of two non-dependent modules, wherein the non-dependent modules are modules that do not have a direct or indirect dependency relationship in the logical dependency relationship; Repeat the selection, crossover, and mutation operations until the termination condition is met, and output the scheduling sequence with the highest fitness value as the installation scheduling scheme for the data center module.

5. The method according to claim 1, characterized in that, Also includes: The installation scheduling plan is executed, and the execution status is monitored in real time. The execution status includes at least one of the following: tower crane working status data, module arrival time data, and environmental parameter data. When a disturbance event is detected, dynamic rescheduling is triggered.

6. The method according to claim 5, characterized in that, The step of triggering dynamic rescheduling upon detecting a disturbance event includes: When the delay time of the actual arrival time of a module relative to the planned time, determined based on the module arrival time data, is less than the first delay threshold, or the timeout duration of the hoisting operation of a single module is less than the first timeout threshold, a local adjustment is triggered. The local adjustment refers to modifying only the adjacent positions of the modules affected by the disturbance in the scheduling sequence. When the delay duration is greater than the first delay threshold and less than the second delay threshold, or when a temporary fault of a single tower crane is detected based on the tower crane's working status data, a regional rescheduling is triggered. The regional rescheduling refers to re-optimizing the scheduling of all modules that have not yet been hoisted within the target area affected by the disturbance; wherein, the first delay threshold is less than the second delay threshold. When the delay duration exceeds the second delay threshold, or when multiple tower cranes are detected to be malfunctioning simultaneously based on the tower crane working status data, or when the environmental parameter data does not meet the preset environmental data, a global rescheduling is triggered. The global rescheduling refers to the global optimization of all modules that have not yet been hoisted.

7. The method according to claim 6, characterized in that, The triggering region rescheduling or global rescheduling includes: Lock the scheduling sequence of completed modules and extract the set of unexecuted modules; Update the scheduling constraints of the set of unexecuted modules; For the set of unexecuted modules, an initial scheduling scheme set is generated based on the logical dependencies, and a new scheduling scheme is generated by global optimization using a multi-objective genetic algorithm with the initial scheduling scheme set as the initial population. The scheduling sequence of the completed modules is merged with the new scheduling scheme to output a complete installation scheduling scheme.

8. A data center module hoisting and scheduling device, characterized in that, include: The first processing module is used to receive module information of the data center module to be installed, wherein the data center module includes at least one of a power supply module, a cooling module, and an information technology equipment module; The second processing module is used to receive scheduling constraints, which include module parameters, equipment parameters, site parameters, and time constraints of the data center modules; and to determine the logical dependencies between the data center modules based on the scheduling constraints. The third processing module is used to generate an initial scheduling scheme set based on the logical dependencies, wherein the initial scheduling scheme set includes multiple scheduling sequences for installing the data center module; The fourth processing module is used to take the initial scheduling scheme set as the initial population, perform global optimization on the scheduling sequence in the initial population through a multi-objective genetic algorithm, and output the installation scheduling scheme of the data center module.

9. An electronic device, characterized in that, include: processor; Memory used to store the processor's executable instructions; The processor is configured to execute the instructions to implement the data center module hoisting scheduling method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores at least one computer program, which is loaded and executed by a processor to enable the computer to implement the data center module hoisting scheduling method as described in any one of claims 1 to 7.

Images