An underwater data center intelligent scheduling method and system based on a water-saving ship lock

Abstract

The application discloses an underwater data center intelligent scheduling method and system based on a water-saving ship lock, which directly utilizes the circulating water flow of the water-saving ship lock to realize natural cooling by deploying the underwater data center in a water delivery channel between a water-saving pool and a lock chamber of the water-saving ship lock. The method comprises the following steps: synchronously collecting ship lock hydraulic, data center cooling and environmental data, and accessing ship and external prediction information; based on data, a discrete time dynamic model of the water-saving pool water volume and the data center inlet water temperature is established; a model predictive control framework is adopted to construct and solve a limited time domain optimization problem taking ship scheduling and cooling water flow as decision variables, taking the minimum comprehensive cost of navigation, heat dissipation and energy consumption as the target and meeting multiple safety constraints, to realize forward-looking and collaborative scheduling and rolling execution; finally, under the premise of ensuring the safety of ship lock operation and equipment heat dissipation, the goal of synergistic effect of water resources, energy and computing power resources is achieved.

Description

Technical Field

[0001] This invention relates to the interdisciplinary field of water conservancy engineering, green energy and information infrastructure, and more specifically, to an intelligent scheduling method and system for underwater data centers based on water-saving ship locks. Background Technology

[0002] A water-saving lock is a water-saving navigation structure that achieves water recycling within the lock chamber through a water-saving pool system. Its basic principle is: when downstream vessels pass through the lock to release water, a portion of the water is stored in the water-saving pool; when upstream vessels pass through the lock to fill water, the water stored in the water-saving pool is used preferentially for replenishment, thus significantly reducing the amount of water drawn from upstream sources and discharged downstream during each lock passage operation. This technology has been successfully applied in several key hubs of inland waterway navigation in my country, demonstrating significant water-saving benefits.

[0003] On the other hand, underwater data centers, as an emerging green computing infrastructure, deploy IT equipment such as servers in sealed underwater hulls and directly utilize the stable low-temperature environment and high heat capacity of water for natural cooling. This can significantly reduce the huge energy consumption caused by cooling in traditional land-based data centers, and has significant advantages in energy saving, water saving, and land saving.

[0004] However, water-saving ship locks and underwater data centers are currently two independently developed technological systems: Functional Independence: The core objective of water-saving locks is to ensure safe and efficient navigation for ships while conserving water resources. Their scheduling algorithms primarily focus on reducing ship waiting times and optimizing gate opening and closing sequences. Underwater data centers, on the other hand, concentrate on the stable operation and efficient heat dissipation of IT equipment, and their cooling control is typically independent of external hydraulic structures.

[0005] Lack of resource synergy: The periodic and large-scale water flow generated during the operation of water-saving locks contains enormous cooling resources, but these are not being effectively utilized. If data centers are built independently, additional cooling water sources or complex heat exchange systems will be required.

[0006] Lack of systematic optimization: When two types of facilities are physically adjacent or combined, they create competing demands for the same water resources: lock navigation requires specific timing of water volume adjustments to raise and lower the lock chamber water level, while data center cooling requires a stable and sufficient flow of cooling water. Current technology lacks a scheduling method and system that can coordinate these two demands and achieve globally optimal scheduling within limited water resources and time windows.

[0007] Therefore, existing technologies suffer from several problems, including underutilization of water energy (cooling capacity), an inability to intelligently reconcile the competition between navigation and heat dissipation, and low overall system energy efficiency. An innovative technological solution is urgently needed that can deeply integrate underwater data centers with water-saving locks, and achieve a synergistic win-win situation of "navigation, heat dissipation, and energy saving" through intelligent scheduling. Summary of the Invention

[0008] The purpose of this invention is to provide an intelligent scheduling method and system for underwater data centers based on water-saving locks, in order to solve the above-mentioned problems existing in the prior art.

[0009] The application is as follows: In a first aspect, the present invention provides an intelligent scheduling method for an underwater data center based on a water-saving lock. The underwater data center is deployed in the water conveyance channel between the water-saving pool and the lock chamber of the water-saving lock. The method is executed in each scheduling cycle and includes the following steps: S1, Multi-source data collection steps: Simultaneously acquire hydraulic status monitoring data of the water-saving lock, cooling system status monitoring data of the underwater data center, related water environment status monitoring data, ship traffic status information, and external boundary condition prediction data. S2, Dynamic Modeling Steps of Coupled System: Based on the real-time data collected in S1, establish a discrete-time mathematical model to describe the dynamic balance of water volume in the water-saving pool and the dynamic change of inlet water temperature in the cooling system of the underwater data center. S3, Collaborative Optimization Decision and Scheduling Steps: Based on the discrete-time mathematical model established in S2, a finite-time domain optimization problem is constructed and solved using a model predictive control framework. The decision variables include ship scheduling sequences and cooling water flow allocation sequences for multiple future scheduling cycles. The optimization objective is to minimize the comprehensive cost that balances navigation efficiency, heat dissipation performance, and control energy consumption. The solution must satisfy multiple constraints, including cooling water flow constraints, water-saving pool water volume constraints, heat dissipation safety constraints, and lock chamber water level safety constraints. The current-time command from the optimal decision sequence obtained is then sent to the water-saving lock control system and the underwater data center cooling system for execution.

[0010] Furthermore, in S1, The hydraulic state monitoring data includes at least the gate chamber water level h. lock (t), water level of the provincial water tank h save (t); The cooling system status monitoring data includes at least the cooling water distribution flow rate Q. c (t), Inlet water temperature of underwater data center cooling system T in_dc (t) and thermal power consumption P of underwater data center heat (t); The associated water body environmental status monitoring data is the ambient water temperature T in the water conveyance channel. env (t); The vessel traffic status information includes at least a list of vessels waiting to pass through the lock, with each vessel's arrival time t. arr,i and priority identifier P s,i ; The external boundary condition prediction data includes at least the upstream inflow forecast Q. inflow (t), downstream water level forecast h downstream (t) and water source temperature trend forecast T source (t).

[0011] Furthermore, the model used to describe the dynamic balance of water volume in the water-saving pool is as follows:

[0012] Among them, V save (t) represents the water storage capacity of the water-saving pool, and the water level h of the water-saving pool is... save (t) Calculated using the pool volume curve; Q in (t) and Q out (t) represents the net flow rate into and out of the water-saving pool at time t, respectively. This is determined by monitoring the water level h in the water-saving pool. save The rate of change of (t) is obtained by combining its cross-sectional area or volume characteristics.

[0013] Furthermore, the model used to describe the dynamic changes in the inlet water temperature of the underwater data center cooling system is as follows:

[0014] Where ρ is the density of water; c w It is the specific heat capacity of water, V dc The effective heat capacity volume of the water in the cooling system is a fixed design value.

[0015] Furthermore, in S3, the process of constructing the optimization objective is as follows: A comprehensive cost function J, in the form of a multi-objective weighted summation, is formed by comprehensively considering the ship waiting time cost representing the navigation efficiency, the inlet water temperature tracking deviation representing the heat dissipation performance, and the cooling water flow rate adjustment representing the energy consumption control, for minimization:

[0016] Where N is the prediction time domain length; The first term α·W s (t+k) is used to characterize the navigation efficiency, W s (t) represents the cost of vessel waiting time, calculated based on the vessel traffic state information obtained in step S1:

[0017] in, For the current waiting ship assembly; The second term β·(T) in_dc (t+k) T ref )² is used to characterize the heat dissipation performance, T ref The target value for the inlet water temperature of the underwater data center cooling system is dynamically set based on the current ambient water temperature in the water supply channel and the preset safety margin ΔT. safe Sure:

[0018] The third term γ·(Q) c (t+k)² is used to characterize the control energy consumption; α, β, and γ are the adjustable weighting coefficients corresponding to the three performance indicators mentioned above.

[0019] Furthermore, when constructing the optimization problem of the model predictive control in S3, the predicted data of the external boundary conditions are injected as feedforward information, specifically for: (a) Based on the water source temperature trend forecast T source (t), determine the predicted sequence {T} of the future water temperature of the water conveyance channel in the model predictive control. env (t+k)}, this sequence directly affects the prediction calculation of the underwater data center cooling system inlet water temperature tracking deviation term, which characterizes heat dissipation performance in the optimization objective; (b) Based on the upstream inflow forecast Q inflow (t), correct the model parameters in the optimization problem that are in dynamic balance with the water volume of the water-saving pool, so as to reflect the expected changes in future available water resources; (c) Based on the downstream water level forecast h downstream (t), determine the time parameter related to the time consumption of the water-saving lock discharge operation; this parameter, as part of the process time constraint, is used to limit the minimum time interval between adjacent operations in the ship scheduling sequence, and then, through the collaborative optimization mechanism of the model predictive control, affect the joint optimal solution of the cooling water flow allocation sequence and the ship scheduling sequence.

[0020] Furthermore, the values ​​of the weighting coefficients α, β, and γ are dynamically configured according to the system's dominant operational requirements, including: When the number of ships waiting to pass through the lock exceeds a preset ship queue length threshold, the value of α is increased; When the thermal power consumption of the underwater data center is higher than a preset percentage of its rated power consumption, or when the difference between the inlet water temperature and the preset maximum allowable water temperature is less than the preset temperature safety margin, the value of β is increased. When both the number of ships waiting to pass through the lock and the thermal power consumption of the underwater data center are lower than their corresponding preset thresholds, the value of γ is increased.

[0021] Furthermore, in S3, the multiple constraints are determined by the system's physical limits and operational safety procedures, and serve as the solution boundary for the optimization problem. These constraints explicitly include at least the following: (1) Cooling water flow constraint:

[0022] Among them, Q c_min With Q c_max These are the minimum and maximum allowable operating flow rates of the cooling water system, respectively. (2) Water volume constraints in the water-saving pool:

[0023] Among them, V save_min With V save_max These are the minimum and maximum water storage limits preset to ensure the safe operation of the water-saving pool; (3) Heat dissipation safety constraints, wherein when the heat dissipation of the underwater data center exceeds a preset ratio of its rated power consumption, the inlet water temperature T of the underwater data center cooling system shall be [not specified]. in_dc (t+k) must satisfy:

[0024] Among them, T max The maximum permissible inlet water temperature to ensure the safety of the underwater data center equipment; (4) Safety constraints on the water level in the sluice chamber:

[0025] Among them, h lock_min with h lock_max These are the minimum and maximum water levels in the lock chamber that are permitted to ensure the safe passage of ships through the lock.

[0026] Secondly, the present invention provides an intelligent scheduling system for an underwater data center based on a water-saving ship lock, comprising: The data acquisition module is deployed in the water-saving lock and underwater data center to collect real-time hydraulic status data of the water-saving lock, cooling system status data of the underwater data center, and related water environment status data. The data access module is used to access ship traffic status information and external boundary condition prediction data; The data processing and decision-making module is communicatively connected to the data acquisition module and the data access module. It is used to establish a dynamic model of the coupled system based on the acquired and accessed data, and solve the collaborative optimization problem to generate water-saving lock operation instructions and cooling water flow regulation instructions. The instruction execution module, connected to the data processing and decision-making module, is used to receive and execute the water-saving lock operation instruction and the cooling water flow regulation instruction.

[0027] Thirdly, the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement an intelligent scheduling method for an underwater data center based on a water-saving lock.

[0028] Compared with the prior art, the embodiments of the present invention achieve the following beneficial effects: By directly embedding the data center into the water conveyance channel of the water-saving lock, the circulating water flow of the water-saving lock can be used as a cooling source for the data center, resulting in significant energy-saving and consumption-reducing economic benefits. Simultaneously, the intelligent collaborative scheduling method based on model predictive control proposed in this invention can proactively integrate multi-source information such as ship traffic, data center load, and hydrological and meteorological forecasts. It dynamically coordinates the competition between navigation water and cooling water in terms of timing and quantity, achieving a globally optimal balance between navigation efficiency, heat dissipation performance, and system energy consumption while strictly ensuring the operational safety of the water-saving lock and the heat dissipation safety of the data center. This improves the efficiency of water resource and energy synergy while ensuring the reliability and intelligent adaptive capability of the dual systems. Furthermore, this solution is based on the modification of existing lock structures, making engineering implementation convenient and cost-effective. It has the potential for large-scale promotion in inland waterway hubs, upgrading traditional water conservancy facilities into a new type of green infrastructure integrating "water-energy-computing," demonstrating significant comprehensive value and strategic importance. Attached Figure Description

[0029] Figure 1 This is a flowchart illustrating an intelligent scheduling method for an underwater data center based on a water-saving lock, provided by an embodiment of the present invention. Detailed Implementation

[0030] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments given herein are for illustration and explanation only and are not intended to limit the present invention.

[0031] It should be noted that many specific details are set forth in the following description in order to provide a full understanding of the present invention. However, the present invention may have other embodiments, and therefore the scope of protection of the present invention is not limited to the specific embodiments disclosed below.

[0032] See attached document Figure 1As shown, this embodiment provides See attached document Figure 1 As shown in the figure, this embodiment provides an intelligent scheduling method for underwater data centers based on water-saving locks. The physical basis of this method is to pre-deploy a modular underwater data center sealed cabin in the water conveyance channel between the water-saving pool and the lock chamber of the water-saving lock, so that its heat dissipation surface is in direct contact with the circulating water in the channel.

[0033] This method is executed on a rolling basis with a fixed scheduling period (e.g., 15 minutes), and each period includes the following three steps: S1. Multi-source data aggregation: At the beginning of each scheduling cycle, the following five types of data are synchronously collected to form a snapshot describing the complete state of the system: 1. Hydraulic status monitoring data of the water-saving ship lock: including at least the lock chamber water level h measured in real time by a water level gauge. lock (t) and the water level h of the provincial reservoir save (t).

[0034] 2. Underwater data center cooling system status monitoring data: including at least the actual cooling water distribution flow rate Q measured by a flow meter. c (t) Inlet water temperature T of the underwater data center cooling system measured by a temperature sensor in_dc (t), and the thermal power P of the underwater data center obtained through a power meter or device management system. heat (t).

[0035] 3. Related water environment status monitoring data: mainly the ambient water temperature T in the water conveyance channel obtained through a temperature measurement point set upstream of the data center cooling water intake. env (t).

[0036] 4. Vessel Traffic Status Information: A list of vessels waiting to pass through the locks is obtained by accessing the lock scheduling system or the Automatic Identification System (AIS). Each record in this list must contain at least the vessel's arrival time t. arr,i and the priority indicator P representing the urgency of its passage through the gate. s,i .

[0037] 5. External boundary condition prediction data: Forecast data received from hydrological and meteorological departments through data interfaces, including at least the upstream inflow forecast Q for the next period of time. inflow (t), downstream water level forecast h downstream (t) and water source temperature trend forecast T source (t).

[0038] S2. Dynamic modeling of coupled systems: Based on real-time data collected by S1, a discrete-time mathematical model describing the core dynamics of the "lock-data center" coupled system is established and maintained. This model comprises two sub-models that are interrelated in terms of water volume and energy, with their state variables being the water storage capacity V of the water-saving reservoir. save (t) and data center inlet water temperature T in_dc (t).

[0039] 1. Dynamic Balance Model of Water Saving Pool: This model describes the conservation relationship of the water saving pool as the water hub of the system, and its expression is:

[0040] Among them, V save (t) represents the water storage capacity of the water-saving pool at time t, and its value is determined by the monitored water level h. save (t) is obtained by conversion using a pre-calibrated pool volume curve. Q in (t) and Q out (t) represents the net flow rate into and out of the provincial water tank at time t, respectively. These two values ​​are not measured directly, but rather through high-frequency monitoring of h. save The rate of change of (t) is obtained in real time by combining the cross-sectional area or volume characteristics of the water-saving pool.

[0041] This equation connects ship scheduling (determining Q) in / Q out ) and cooling control (determines Q) c The core physical link is that water resources used for cooling and water resources used for navigation share the total volume V of the water-saving pool. save There is a competitive relationship among them.

[0042] 2. Dynamic Model of Inlet Water Temperature for Underwater Data Center Cooling System: This model, based on the law of conservation of energy, describes how cooling water flow rate affects the inlet water temperature of the data center. Its expression is as follows:

[0043] Where ρ is the density of water, and c w It is the specific heat capacity of water, V dc The effective heat capacity volume of the water in the internal pipes and heat exchangers of the data center cooling system is a fixed value determined by the system design.

[0044] The equation clearly reveals the control mechanism. The first term within the curly braces on the right... This represents the amount of heat (cooling) that the cooling water carries away from the system per unit time, and its magnitude is directly proportional to the flow rate. and temperature difference ( ). Second item This represents the heat generated within the system. The difference between the two is the net heat gain of the system, divided by the heat capacity. Then, multiplying by the time step Δt, we obtain the change in inlet water temperature for the next cycle. Therefore, by adjusting... It can dynamically balance changes. Thus It remains stable within the target range.

[0045] S3. Collaborative Optimization Decision-Making and Scheduling Based on the discrete-time mathematical model established in S2, a finite-time optimization problem is constructed and solved using a model predictive control framework. Specifically, this includes: S31. Optimization Problem Modeling: At the beginning of each scheduling cycle (time t), an optimization problem for the next N scheduling cycles is constructed. The construction of the optimization problem includes the following key elements: 1. Definition of decision variables: The goal of solving the optimization problem is to determine a set of optimal operation instructions for each future scheduling cycle, which is specifically represented by two sequences of decision variables: a. Ship scheduling sequence {u s (t|t),...,u s (t+N-1|t)}: This sequence is a discrete decision set, where each u s (t+k|t) defines the specific operation that the lock chamber will perform during the (t+k)th scheduling period, such as "perform downstream water discharge operation for ships A and B in the waiting queue" or "remain idle". This decision is directly related to the filling and emptying valve operation of the lock and the order of ship entry and exit.

[0046] b. Cooling water flow distribution sequence {Q} c (t|t),...,Q c (t+N-1|t)}: This sequence is a continuous decision set, where each Q... c (t+k|t) defines the water flow rate that is planned to be allocated to the underwater data center for cooling during the t+kth scheduling period.

[0047] 2. Optimization Objective Function Construction: Construct a comprehensive cost function J as the objective of the optimization problem, i.e., the object to be minimized. This function unifies the three sub-objectives of navigation efficiency, heat dissipation performance, and control energy consumption into a unified mathematical framework for quantitative trade-offs. Its specific form is the following multi-objective weighted summation:

[0048] The first term α·W s (t+k) is used to characterize the navigation efficiency, W s (t) represents the cost of vessel waiting time, calculated based on the vessel traffic state information obtained in step S1:

[0049] in, For the current waiting ship assembly; The second term β·(T) in_dc (t+k) T ref )² is used to characterize the heat dissipation performance, T ref The target value for the inlet water temperature of the underwater data center cooling system is dynamically set based on the current ambient water temperature in the water supply channel and the preset safety margin ΔT. safe Sure:

[0050] The third term γ·(Q) c (t+k)² is used to characterize the control energy consumption; α, β, and γ are the adjustable weighting coefficients corresponding to the three performance indicators mentioned above.

[0051] 3. Forward-looking correction: When constructing the above optimization problem, the predicted external boundary conditions obtained in step S1 are injected as feedforward information to dynamically correct the problem and achieve forward-looking optimization. a. Water temperature prediction correction: based on T source (t) Forecast, determine the model predictive control internals for the future T env The predicted sequence of (t+k) allows the heat dissipation decision to adapt to changes in water source temperature in advance, so as to more accurately predict the heat dissipation effect.

[0052] b. Water volume forecast correction: based on Q inflow (t) forecast, corrects the prediction parameters in the dynamic water balance model, and reflects the expected changes in future available water resources.

[0053] c. Time constraint correction: based on h downstream (t) Forecast and dynamically calculate the time parameter for the water release operation. This parameter is used to limit the minimum time interval between adjacent operations in the ship scheduling sequence, thereby affecting the ship scheduling sequence {u s The feasibility of (t+k)}.

[0054] 4. Multiple constraints: The solution must satisfy the following boundary conditions determined by physical limits and safety regulations: System dynamic constraints: The prediction of the future system state by the decision variables must conform to the discrete-time mathematical model established in the S2 step, namely the dynamic balance model of the water volume in the water-saving pool and the dynamic model of the inlet water temperature of the underwater data center cooling system. This is the fundamental physical constraint of the optimization problem.

[0055] Cooling water flow constraints: The cooling water flow allocated to the data center must be within the capacity of the executing equipment, i.e., the inequality must be satisfied:

[0056] Among them, Q c_min With Q c_max These are the minimum and maximum allowable operating flow rates of the cooling water system, respectively.

[0057] Water-saving tank water volume constraint: The water storage capacity of the water-saving tank must be maintained within its safe operating range to ensure continuous system operation, which means the following inequality must be satisfied:

[0058] V save_min With V save_max These are the minimum and maximum water storage limits preset to ensure the safe operation of the water-saving pool.

[0059] Thermal safety constraints: To ensure the safety of underwater data center equipment, when operating under high load, the inlet water temperature of its cooling system must not exceed the safety limit, i.e., it must meet the following requirement: when the thermal power consumption P of the underwater data center... heat When (t+k) exceeds a preset percentage of its rated power consumption, the following is mandatory:

[0060] Among them, T max The maximum permissible inlet water temperature to ensure the safety of the underwater data center equipment.

[0061] Lock chamber water level safety constraints: The lock chamber water level must be maintained within the allowable range to ensure the safe passage of ships, which means that the following inequality must be satisfied:

[0062] Among them, h lock_min with h lock_max These are the minimum and maximum water levels in the lock chamber that are permitted to ensure the safe passage of ships through the lock.

[0063] Lock operation logic and safety regulations: determined by ship scheduling decisions. s The operational rules directly triggered by (t+k) include, but are not limited to: the mutual exclusion of lock chamber filling and discharge operations, restrictions on the number and size of vessels passing through the lock at one time, the minimum time requirements for each stage of vessel passage through the lock (entering the lock, closing the gate, adjusting the water level, opening the gate, and exiting the lock), and the bidirectional passage intervals. These constraints limit the feasible space for vessel scheduling sequences.

[0064] S32. Optimization Problem Solving and Instruction Generation: The central processing unit (CPU) combines the objective function, the dynamic model in S2, and all constraints into a mixed-integer optimization problem. It then uses a dedicated solver (such as CPLEX or Gurobi) or an efficient heuristic algorithm (such as a genetic algorithm) to solve the problem online quickly, obtaining the optimal decision sequence {u}. s *,Q c *}

[0065] S33. Rolling Optimization and Execution: Extract only the decision [u] corresponding to the current time t in the optimal sequence s *(t|t),Q c *(t|t)].

[0066] will u s *(t|t) is converted into specific lock operation instructions (e.g., opening valve X to release water into water-saving pool Y; arranging for vessels A and B to enter the lock), and sent to the water-saving lock centralized control system.

[0067] Q c *(t|t) is used as a set value and sent to the flow regulating valve or variable frequency water pump of the underwater data center cooling system.

[0068] At the end of the current cycle, the system state changes due to external disturbances (arrival of new ships, load changes) and its own control actions. At the start of the next scheduling cycle t+1, the system repeats the entire S1→S2→S3 process based on the latest state data collected in S1: that is, updating the model, reconstructing and solving a new finite-time optimization problem, and executing the new optimal command for the current moment. This mechanism of "iterative online optimization, rolling implementation, and feedback correction" is the core of the model predictive control framework, ensuring that the system always maintains a dynamically optimal or suboptimal operating state when facing uncertainties.

[0069] S34. Adaptive Strategy Implementation: To enable the system to intelligently respond to different dominant operational needs, the weight coefficients α, β, and γ in the objective function are dynamically configured based on the real-time operational status. The system presets several operating modes and switches between them according to the following rules: Peak Hour Traffic Flow Mode: When the number of vessels waiting to pass through the lock exceeds the preset vessel queue length threshold, the system determines that traffic pressure is the primary issue. At this time, the system automatically increases the value of the traffic efficiency weight α, while correspondingly decreasing the values ​​of the heat dissipation weight β and the control energy consumption weight γ. In this mode, optimization decisions will prioritize shortening vessel waiting times and quickly diverting traffic.

[0070] Computing power protection mode: When the thermal power consumption P of the underwater data center is detected... heat (t) exceeds a preset percentage of its rated power consumption, or its cooling system inlet water temperature Tin_dc (t) and the preset maximum allowable water temperature T max When the difference between the cooling performance weight β and the energy consumption weight γ is less than the preset temperature safety margin, the system determines that there is a risk to heat dissipation safety. In this case, the system automatically increases the value of the heat dissipation performance weight β, while correspondingly decreasing the values ​​of the navigability weight α and the energy consumption weight γ. Under this mode, optimization decisions will prioritize ensuring the cooling effect of the data center to guarantee stable computing power.

[0071] Green and energy-saving mode: When the number of ships waiting to pass through the lock is monitored and the thermal power consumption P of the underwater data center is measured... heat When both (t) are below their corresponding preset thresholds, the system determines that both navigation and heat dissipation pressures are relatively low. At this time, the system automatically increases the control energy consumption weight γ, while correspondingly decreasing the navigation weight α and the heat dissipation weight β. In this mode, optimization decisions will tend to smooth control actions, minimize the adjustment energy consumption of actuators such as pumps and valves, and achieve optimal system operating energy efficiency.

[0072] Corresponding to the above method, this embodiment also provides an intelligent scheduling system for an underwater data center based on a water-saving lock. This system is deployed at the water-saving lock and underwater data center site and includes the following functional modules: 1. Data Acquisition Module: Composed of a physical sensor network deployed in the lock chamber, water-saving pool, water conveyance channel, and underwater data center cooling circuit, including water level gauges, temperature sensors, flow meters, power meters, etc., used to collect real-time hydraulic status data of the water-saving lock, status data of the underwater data center cooling system, and related water environment status data.

[0073] 2. Data Access Module: Composed of an industrial communication gateway and a protocol conversion interface, it is used to access ship traffic status information from the lock dispatch center or AIS base station, as well as external boundary condition prediction data from external hydrological and meteorological departments.

[0074] 3. Data Processing and Decision-Making Module: As the "brain" of the system, this module is typically implemented by an industrial control computer or a high-performance edge server. It connects to the data acquisition and data access modules via an industrial Ethernet network. Internally, it runs dedicated control software to execute all the computational and decision-making logic of the aforementioned intelligent scheduling methods. This involves multi-source data aggregation, coupled system dynamic modeling, and collaborative optimization decision-making and scheduling, ultimately generating water-saving lock operation commands and cooling water flow regulation commands.

[0075] 4. Command Execution Module: This module consists of actuators such as the lock gate / valve drive unit, regulating valves or variable frequency pumps in the data center cooling water circuit. It connects to the data processing and decision-making module to receive and execute specific control commands, thereby driving the physical system operation.

[0076] This embodiment also relates to an electronic device. The device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. The processor is configured to, when executing the program, implement all the steps of the aforementioned intelligent scheduling method for underwater data centers.

[0077] Specifically, when executing a program, the processor is configured to automatically perform the following operations in each scheduling cycle: The system synchronously collects monitoring data such as lock hydraulics, data center cooling system status, and water environment through communication interfaces, and also accesses ship traffic information and external forecast data.

[0078] Based on the collected data, a dynamic prediction model for the water volume of the water-saving pool and the inlet water temperature of the data center was established and updated.

[0079] A model predictive control framework is adopted to construct and solve a finite-time domain optimization problem with ship scheduling sequence and cooling water flow sequence as decision variables, minimizing comprehensive operating cost as objective, and satisfying multiple safety constraints.

[0080] The optimal command obtained from the solution is immediately sent to the lock control system and the data center cooling system for execution.

[0081] This electronic device can be deployed in a control center. Its memory is used to store programs and data, and its processor is responsible for performing calculations and logic control, together realizing the automated operation of the intelligent scheduling function.

[0082] Numerous specific details are set forth in the specification provided herein. However, it will be understood that embodiments of the invention may be practiced without these specific details. In some instances, well-known methods, structures, and techniques have not been shown in detail so as not to obscure the understanding of this specification.

[0083] Furthermore, those skilled in the art will understand that although some embodiments herein include certain features included in other embodiments but not others, combinations of features from different embodiments are intended to be within the scope of the invention and form different embodiments. Any of the claimed embodiments can be used in any combination.

Claims

1. A method for intelligent scheduling of underwater data centers based on water-saving ship locks, characterized in that, The underwater data center is deployed in the water conveyance channel between the water-saving pool and the lock chamber of the water-saving lock. The method is executed in each scheduling cycle and includes the following steps: S1. Multi-source data aggregation: Simultaneously acquire hydraulic status monitoring data of the water-saving lock, cooling system status monitoring data of the underwater data center, related water environment status monitoring data, ship traffic status information, and external boundary condition prediction data; S2. Dynamic Modeling of Coupled Systems: Based on the real-time data collected in S1, a discrete-time mathematical model is established to describe the dynamic balance of water volume in the water-saving pool and the dynamic change of inlet water temperature in the cooling system of the underwater data center. S3. Collaborative Optimization Decision and Scheduling: Based on the discrete-time mathematical model established in S2, a model predictive control framework is used to construct and solve a finite-time domain optimization problem. The decision variables include ship scheduling sequences and cooling water flow allocation sequences for multiple future scheduling cycles. The optimization objective is to minimize the comprehensive cost that balances navigation efficiency, heat dissipation performance, and control energy consumption. The solution must satisfy multiple constraints, including cooling water flow constraints, water-saving pool water volume constraints, heat dissipation safety constraints, and lock chamber water level safety constraints. The current-time command in the optimal decision sequence obtained from the solution is sent to the water-saving lock control system and the underwater data center cooling system for execution.

2. The intelligent scheduling method for underwater data centers based on water-saving locks according to claim 1, characterized in that, In S1, The hydraulic state monitoring data includes at least the gate chamber water level h. lock (t), water level of the provincial water tank h save (t); The cooling system status monitoring data includes at least the cooling water distribution flow rate Q. c (t), Inlet water temperature of underwater data center cooling system T in_dc (t) and thermal power consumption P of underwater data center heat (t); The associated water body environmental status monitoring data is the ambient water temperature T in the water conveyance channel. env (t); The vessel traffic status information includes at least a list of vessels waiting to pass through the lock, with each vessel's arrival time t. arr,i and priority identifier P s,i ; The external boundary condition prediction data includes at least the upstream inflow forecast Q. inflow (t), downstream water level forecast h downstream (t) and water source temperature trend forecast T source (t).

3. The intelligent scheduling method for underwater data centers based on water-saving locks according to claim 1, characterized in that, The model used to describe the dynamic balance of water volume in the water-saving pool is as follows: Among them, V save (t) represents the water storage capacity of the water-saving pool, and the water level h of the water-saving pool is... save (t) Calculated using the pool volume curve; Q in (t) and Q out (t) represents the net flow rate into and out of the water-saving pool at time t, respectively. This is determined by monitoring the water level h in the water-saving pool. save The rate of change of (t) is obtained by combining its cross-sectional area or volume characteristics.

4. The intelligent scheduling method for underwater data centers based on water-saving locks according to claim 1, characterized in that, The model used to describe the dynamic changes in the inlet water temperature of the underwater data center cooling system is as follows: Where ρ is the density of water; c w It is the specific heat capacity of water, V dc The effective heat capacity volume of the water in the cooling system is a fixed design value.

5. The intelligent scheduling method for underwater data centers based on water-saving locks according to claim 1, characterized in that, In S3, the process of constructing the optimization objective is as follows: The comprehensive cost function J, which combines the ship waiting time cost representing the navigation efficiency, the inlet water temperature tracking deviation representing the heat dissipation performance, and the cooling water flow rate adjustment representing the energy consumption control, is formed into a multi-objective weighted summation form for minimization: Where N is the prediction time domain length; The first term α·W s (t+k) is used to characterize the navigation efficiency, W s (t) represents the cost of ship waiting time, calculated based on the ship traffic state information obtained in step S1: in, For the current waiting ship assembly; The second term β·(T) in_dc (t+k) T ref )² is used to characterize the heat dissipation performance, T ref The target value for the inlet water temperature of the underwater data center cooling system is dynamically set based on the current ambient water temperature in the water supply channel and the preset safety margin ΔT. safe Sure: The third term γ·(Q) c (t+k)² is used to characterize the control energy consumption; α, β, and γ are the adjustable weighting coefficients corresponding to the three performance indicators mentioned above.

6. The intelligent scheduling method for underwater data centers based on water-saving locks according to claim 1, characterized in that, When constructing the optimization problem of the model predictive control in S3, the predicted data of the external boundary conditions are injected as feedforward information, specifically for: (a) Based on the water source temperature trend forecast T source (t), determine the predicted sequence {T} of the future water temperature of the water conveyance channel in the model predictive control. env (t+k)}, this sequence directly affects the prediction calculation of the underwater data center cooling system inlet water temperature tracking deviation term, which characterizes heat dissipation performance in the optimization objective; (b) Based on the upstream inflow forecast Q inflow (t), modify the model parameters in the optimization problem that are in dynamic balance with the water volume of the water-saving pool, so as to reflect the expected changes in future available water resources; (c) Based on the downstream water level forecast h downstream (t), determine the time parameter related to the time consumption of the water-saving lock discharge operation; this parameter, as part of the process time constraint, is used to limit the minimum time interval between adjacent operations in the ship scheduling sequence, and then, through the collaborative optimization mechanism of the model predictive control, affect the joint optimal solution of the cooling water flow allocation sequence and the ship scheduling sequence.

7. The intelligent scheduling method for underwater data centers based on water-saving locks according to claim 5, characterized in that, The values ​​of the weighting coefficients α, β, and γ are dynamically configured according to the system's primary operational requirements, including: When the number of ships waiting to pass through the lock exceeds a preset ship queue length threshold, the value of α is increased; When the thermal power consumption of the underwater data center is higher than a preset percentage of its rated power consumption, or when the difference between the inlet water temperature and the preset maximum allowable water temperature is less than the preset temperature safety margin, the value of β is increased. When both the number of ships waiting to pass through the lock and the thermal power consumption of the underwater data center are lower than their corresponding preset thresholds, the value of γ is increased.

8. The intelligent scheduling method for underwater data centers based on water-saving locks according to claim 1, characterized in that, In S3, the multiple constraints are determined by the system's physical limits and operational safety procedures, and serve as the solution boundary for the optimization problem. These constraints explicitly include at least the following: (1) Cooling water flow constraint: Among them, Q c_min With Q c_max These are the minimum and maximum allowable operating flow rates of the cooling water system, respectively. (2) Water volume constraints in the water-saving pool: Among them, V save_min With V save_max These are the minimum and maximum water storage limits preset to ensure the safe operation of the water-saving pool; (3) Heat dissipation safety constraints, wherein when the heat dissipation of the underwater data center exceeds a preset ratio of its rated power consumption, the inlet water temperature T of the underwater data center cooling system shall be [not specified]. in_dc (t+k) must satisfy: Among them, T max The maximum permissible inlet water temperature to ensure the safety of the underwater data center equipment; (4) Safety constraints on the water level in the sluice chamber: Among them, h lock_min with h lock_max These are the minimum and maximum water levels in the lock chamber that are permitted to ensure the safe passage of ships through the lock.

9. An intelligent scheduling system for an underwater data center based on a water-saving ship lock, used to implement the method as described in any one of claims 1 to 8, characterized in that, include: The data acquisition module is deployed in the water-saving lock and underwater data center to collect real-time hydraulic status data of the water-saving lock, cooling system status data of the underwater data center, and related water environment status data. The data access module is used to access ship traffic status information and external boundary condition prediction data; The data processing and decision-making module is communicatively connected to the data acquisition module and the data access module. It is used to establish a dynamic model of the coupled system based on the acquired and accessed data, and solve the collaborative optimization problem to generate water-saving lock operation instructions and cooling water flow regulation instructions. The instruction execution module, connected to the data processing and decision-making module, is used to receive and execute the water-saving lock operation instruction and the cooling water flow regulation instruction.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the intelligent scheduling method for underwater data centers based on water-saving locks as described in any one of claims 1 to 8.

Images