Regional centralized cooling collaborative control method based on electric refrigeration and dynamic ice storage and related equipment

Abstract

The application discloses a regional centralized cooling collaborative control method based on electric refrigeration and dynamic ice cold storage and related equipment, and relates to the technical field of regional centralized cooling. The application discloses a regional centralized cooling collaborative control method based on electric refrigeration and dynamic ice cold storage and related equipment, and relates to the technical field of regional centralized cooling. The method comprises the following steps: predicting future 24-hour hourly cooling load through convolution, bidirectional long short-term memory neural network and multi-head attention mechanism; acquiring a future 24-hour day-ahead cold source output plan; acquiring a future 4-hour cold source output set value sequence; adjusting pipe network water pump frequency and valve opening degree through feedforward compensation and PID according to the cold source output set value sequence; controlling ice cold storage in response to the day-ahead cold source output plan and the hourly cooling load; balancing pipe network water force through a distributed model predictive control strategy in response to the hourly cooling load and pipe network operation state data; adjusting water supply temperature or cooling power for users of different types; and adjusting the power consumption of the regional cooling system in response to the hourly cooling load and equipment operation parameters. The application can improve cold storage efficiency and cold release precision and can be widely applied to the technical field of regional centralized cooling.

Description

Technical Field

[0001] This invention relates to the field of district cooling technology, and in particular to a district cooling coordinated control method and related equipment based on electric refrigeration and dynamic ice storage. Background Technology

[0002] District cooling refers to a power supply method that targets a group of buildings within a certain area, producing chilled water or cooling capacity at a centralized energy station and distributing it to individual users through underground pipelines to meet the cooling load demands of various buildings within the area. Ice storage technology is a widely used cooling capacity storage technology in district cooling systems. It utilizes off-peak electricity at night to produce ice and store cooling capacity, then melts the ice during peak daytime electricity demand to provide cooling, thus achieving peak-shifting and valley-filling of electricity consumption.

[0003] In recent years, dynamic ice slurry cold storage technology has gradually emerged. Compared with traditional static ice storage (ice coils or ice balls), dynamic ice storage uses direct heat exchange between refrigerant and water to generate flocculent ice crystals, resulting in higher energy efficiency, faster load response, and a reduction in floor space of over 30%. Subcooled water dynamic ice-making technology produces -2℃ subcooled water through a plate heat exchanger, which is then used to generate ice slurry after crystallization, achieving high-efficiency heat transfer due to the low temperature difference. However, the systematic application of dynamic ice slurry cold storage in district cooling systems remains unexplored. Summary of the Invention

[0004] In view of this, the main objective of the embodiments of the present invention is to provide a regional centralized cooling coordinated control method and related equipment based on electric refrigeration and dynamic ice storage, in order to solve at least one of the problems of the prior art. The present invention can improve the cold storage efficiency and the cold release accuracy.

[0005] To achieve the above objectives, one aspect of the present invention provides a method for coordinated control of regional centralized cooling based on electric refrigeration and dynamic ice storage, the method comprising: Dynamically collect meteorological data, building operation data, user behavior data, pipeline operation status data, and equipment operation parameters; Based on the meteorological data, the building operation data, and the user behavior data, the hourly cooling load for the next 24 hours is predicted using a convolutional neural network, a bidirectional long short-term memory neural network, and a multi-head attention mechanism. Based on the time-of-use electricity pricing mechanism, obtain the day-ahead cooling source output plan for the next 24 hours; Based on the pipeline operation status data and equipment operation parameters, dynamically obtain the cold source output setpoint sequence for the next 4 hours; Based on the pipeline network operation status data and the setpoint sequence of the cold source output for the next 4 hours, the frequency of the pipeline pumps and the valve opening are finely adjusted through feedforward compensation and PID algorithm. The ice storage is dynamically controlled in response to the daytime cooling source output plan for the next 24 hours and the hourly cooling load for the next 24 hours. In response to the hourly cooling load for the next 24 hours and the pipeline network operation status data, a distributed model predictive control strategy is used to dynamically balance the hydraulic system of the pipeline network. Based on the building operation data, users of various business types are classified, and the water supply temperature or cooling power is dynamically adjusted according to the classification results. The power consumption of the district cooling system is adjusted in response to the hourly cooling load for the next 24 hours and the equipment operating parameters.

[0006] In some embodiments, predicting the hourly cooling load for the next 24 hours based on the meteorological data, the building operation data, and the user behavior data, using a convolutional neural network, a bidirectional long short-term memory neural network, and a multi-head attention mechanism, includes the following steps: Based on the meteorological data, the building operation data, and the user behavior data, the cold load sequence of each of the aforementioned business types is decomposed to obtain load components; A hybrid model is constructed based on the convolutional neural network, the bidirectional long short-term memory neural network, and the multi-head attention mechanism; The hybrid model is used to predict each load component, and the prediction results are obtained. The prediction results are superimposed to obtain the hourly cooling load for the next 24 hours.

[0007] In some embodiments, obtaining the day-ahead cooling source output plan for the next 24 hours based on the time-of-use electricity pricing mechanism includes the following steps: Based on the pipeline length, thermal resistance of the insulation layer, and supply and return water temperatures, a cooling loss term is constructed. Based on the cooling loss term and the total operating cost of the regional centralized cooling system, construct the objective function; The output range of the electric refrigeration unit, the cold storage and release rate of the dynamic ice storage device, the upper and lower limits of the cold storage capacity of the ice storage tank, and the constraints on the reliability of the cooling supply are used as limiting conditions. Based on the objective function and the constraints, a mixed-integer linear programming approach is used to obtain the daytime cooling source output plan for the next 24 hours. The daytime cold source output plan includes the start / stop status and cooling power setting of each electric refrigeration unit, the ice storage capacity plan of the dynamic ice storage device, and the ice slurry release capacity plan.

[0008] In some embodiments, dynamically obtaining the setpoint sequence of the cooling source output for the next 4 hours based on the pipeline operation status data and equipment operation parameters includes the following steps: Based on the charging and discharging characteristics of the ice storage device, a state-space model is constructed; Based on the pipeline operation status data and equipment operation parameters, the system status of the regional centralized cooling system at each time point in the next 4 hours is recursively predicted through the state space model. Using the state-space model, a mapping relationship between preset physical constraints and the system state is established, and a quadratic programming problem is constructed based on the mapping relationship. The solution to the quadratic programming problem is obtained, and the setpoint sequence of the cold source output for the next 4 hours is obtained.

[0009] In some embodiments, the step of fine-tuning the frequency of the water pumps and the valve opening based on the pipeline network operating status data and the setpoint sequence of the cold source output for the next 4 hours, through feedforward compensation and PID algorithm, includes the following steps: Linear interpolation is performed on the setpoint sequence of cold source output for the next 4 hours to obtain the predicted value of cold source output change for the next 15 minutes as the feedforward compensation value. The supply and return water temperatures in the pipeline network operation status data are collected once per second. The deviation between the supply and return water temperatures and the set values ​​is obtained, and a feedback correction value is output through a PID algorithm. The feedforward compensation value and the feedback correction value are superimposed to generate an execution command to adjust the frequency of the pipeline pumps and the valve opening.

[0010] In some embodiments, dynamically controlling ice storage in response to the daytime cooling source output plan for the next 24 hours and the hourly cooling load for the next 24 hours includes the following steps: During periods of low electricity prices, in response to the ice storage plan in the daytime cold source output plan for the next 24 hours, the electric refrigeration unit is switched to ice-making mode, and the ultrasonic crystallization unit is activated to reduce the supercooling of ice formation, generating flocculent ice crystals and storing them in the ice slurry storage tank. During peak electricity price periods, in response to the hourly cooling load of the next 24 hours, priority is given to controlling the ice slurry storage tank to release ice slurry for cooling. When the hourly cooling load for the next 24 hours exceeds the ice slurry release capacity, it will automatically switch to a combined cooling mode of electric refrigeration unit and ice slurry release. During the ice slurry release process, the frequency of the ice slurry delivery pump and the valve opening are dynamically adjusted based on the supply and return water temperatures in the pipeline operation status data.

[0011] In some embodiments, the dynamic balance control of the pipeline hydraulics, in response to the hourly cooling load of the next 24 hours and the pipeline operating status data, is performed through a distributed model predictive control strategy, including the following steps: The hourly cooling load for the next 24 hours is used as a feedforward input to adjust the opening of the regulating valves and the frequency of the circulating pumps in each branch of the pipe network. Based on the flow and pressure of each branch and the pressure difference of each heat exchange station in the pipeline network operation status data, the opening of the regulating valves of each branch of the pipeline network is coordinated and adjusted through a distributed model predictive control strategy.

[0012] To achieve the above objectives, another aspect of the present invention proposes a regional centralized cooling coordinated control device based on electric refrigeration and dynamic ice storage, the device comprising: The multi-source data acquisition module is used to dynamically collect meteorological data, building operation data, user behavior data, pipeline operation status data, and equipment operation parameters; The cooling load prediction module is used to predict the hourly cooling load for the next 24 hours based on the meteorological data, the building operation data, and the user behavior data, using a convolutional neural network, a bidirectional long short-term memory neural network, and a multi-head attention mechanism. The day-ahead optimization module is used to obtain the day-ahead cooling source output plan for the next 24 hours based on the time-of-use electricity pricing mechanism; The intraday rolling correction module is used to dynamically obtain the sequence of cold source output setpoints for the next 4 hours based on the pipeline operation status data and equipment operation parameters. The real-time correction module is used to fine-tune the frequency of the water pumps and the valve opening based on the pipeline network operation status data and the set value sequence of the cold source output for the next 4 hours, through feedforward compensation and PID algorithm. The dynamic ice storage intelligent scheduling module is used to dynamically control ice storage in response to the daytime cold source output plan for the next 24 hours and the hourly cooling load for the next 24 hours. The pipeline hydraulic balance control module is used to dynamically balance the pipeline hydraulics in response to the hourly cooling load and pipeline operation status data for the next 24 hours, through a distributed model predictive control strategy. The user load response and differentiated scheduling module is used to classify users of various business types according to the building operation data, and dynamically adjust the water supply temperature or cooling power according to the classification results. The virtual power plant collaborative control module is used to adjust the power consumption of the district cooling system in response to the hourly cooling load and the equipment operating parameters for the next 24 hours.

[0013] To achieve the above objectives, another aspect of the present invention provides an electronic device, the electronic device including a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the method described above.

[0014] To achieve the above objectives, another aspect of the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the methods described above.

[0015] To achieve the above objectives, another aspect of the present invention provides a computer program product or computer program that includes computer instructions stored in a computer-readable storage medium. A processor of a computer device can read the computer instructions from the computer-readable storage medium and execute the computer instructions to cause the computer device to perform the aforementioned method.

[0016] The embodiments of the present invention include at least the following beneficial effects: The present invention provides a method and related equipment for coordinated control of regional centralized cooling based on electric refrigeration and dynamic ice storage. This scheme provides a data foundation for subsequent precise regulation by dynamically collecting meteorological data, building operation data, user behavior data, pipeline network operation status data, and equipment operation parameters; based on meteorological data, building operation data, and user behavior data, it predicts the hourly cooling load for the next 24 hours through convolutional neural networks, bidirectional long short-term memory neural networks, and multi-head attention mechanisms, providing uncertainty boundaries for optimized scheduling; based on the time-of-use electricity pricing mechanism, it obtains the day-ahead cooling source output plan for the next 24 hours, significantly reducing the overall operating cost of the regional centralized cooling system while meeting rigid cooling demand; based on pipeline network operation status data and equipment operation parameters, it dynamically obtains the setpoint sequence of cooling source output for the next 4 hours, enabling real-time correction of deviations in the day-ahead plan; based on pipeline network operation status data... Based on the setpoint sequence of cooling source output for the next 4 hours, feedforward compensation and PID algorithms are used to fine-tune the frequency of water pumps and valve openings in the pipeline network, effectively suppressing hydraulic imbalance in the pipeline network. Responding to the daytime cooling source output plan and hourly cooling load for the next 24 hours, ice storage is dynamically controlled, improving the heat exchange efficiency and service life of the ice storage device, and enhancing storage efficiency and release accuracy. Responding to the hourly cooling load and pipeline operation status data for the next 24 hours, a distributed model predictive control strategy is used to dynamically balance the hydraulic system of the pipeline network, achieving on-demand allocation of cooling capacity and significantly reducing ineffective energy consumption in the transmission and distribution process. Based on building operation data, users of various business types are classified, and the water supply temperature or cooling power is dynamically adjusted according to the classification results, balancing user comfort with grid supply and demand balance. Responding to the hourly cooling load and equipment operating parameters for the next 24 hours, the power consumption of the district cooling system is adjusted, improving the overall stability of the energy system. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of a regional centralized cooling coordinated control method based on electric refrigeration and dynamic ice storage provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of the hardware structure of the electronic device provided in an embodiment of the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with those of this invention; they are merely examples of apparatuses and methods consistent with some aspects of the embodiments of this invention as detailed in the appended claims.

[0020] It should be noted that although functional modules are divided in the system diagram and a logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the system or the order in the flowchart. The terms "first / S100" and "second / S200" in the specification, claims, and the foregoing drawings may be used herein to describe various concepts, but unless specifically stated otherwise, these concepts are not limited by these terms. These terms are used only to distinguish one concept from another. For example, first information may also be referred to as second information without departing from the scope of the embodiments of the invention, and similarly, second information may also be referred to as first information. Depending on the context, the words "if" or "when" as used herein may be interpreted as "when," "in response to a determination," or "in the event of a determination."

[0021] The terms “at least one,” “multiple,” “each,” “any,” etc., used in this invention, “at least one” includes one, two, or more than two; “multiple” includes two or more than two; “each” refers to each of the corresponding multiple; and “any” refers to any one of the multiple.

[0022] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein is for the purpose of describing embodiments of the invention only and is not intended to limit the invention.

[0023] Before providing a detailed description of the embodiments of the present invention, some of the nouns and terms involved in the embodiments of the present invention will be explained first. The nouns and terms involved in the embodiments of the present invention are subject to the following interpretations.

[0024] Water supply temperature refers to the temperature of the chilled water that is prepared by the cooling station (or cold source) and then delivered to each user terminal through the water supply pipeline.

[0025] Return water temperature refers to the temperature at which chilled water, after absorbing heat from the indoor environment through the fan coil unit or air conditioning unit at the user's terminal, flows back to the cold source through the return water pipe. It is usually higher than the supply water temperature.

[0026] Supply and return water temperatures are a collective term for the combination of supply and return water temperatures, referring to this pair of inlet and outlet parameters. In certain specific contexts, it may also refer to the temperature difference between the two, and this temperature difference is directly related to the system's energy consumption and heat exchange efficiency.

[0027] Mixed-integer linear programming (MILP) is a mathematical optimization method that combines the characteristics of linear programming and integer programming. It allows some variables to be integers (discrete decisions) and some variables to be continuous values ​​(partitionable resources), thus enabling the solution of complex practical problems such as production planning, logistics optimization, and investment decisions.

[0028] Model predictive control (MPC) is an advanced control strategy, particularly suitable for multivariable systems and constrained control problems. Its core idea is to achieve dynamic control through predictive models, rolling optimization, and feedback correction.

[0029] Convolutional Neural Networks (CNNs) are a type of deep learning algorithm that is particularly effective in image processing and computer vision. By mimicking the mechanisms of the human visual system, CNNs can automatically learn and extract useful features from images to perform tasks such as image classification, object detection, and image segmentation.

[0030] Bidirectional Long Short-Term Memory (BiLSTM) is a special type of recurrent neural network (RNN) that combines forward and backward information flows to better capture contextual information in sequential data.

[0031] Multi-head self-attention is a core component of the Transformer architecture. By running multiple self-attention heads in parallel, it captures sequence features from different subspaces, significantly improving the model's expressive and generalization capabilities. Its core idea is to project the input vector onto multiple low-dimensional subspaces, independently compute attention in each subspace, and then fuse the results.

[0032] In related technologies, existing district cooling systems based on electric refrigeration and ice storage mostly rely on static historical data or simple time series models to predict the next day's load, making it difficult to capture sudden load changes. Although deep learning (such as CNN-LSTM and attention mechanisms) has been widely used in load forecasting, there is still a lack of effective means to decompose and predict the behavioral characteristics of users in district cooling systems. Existing ice storage systems mostly adopt day-ahead optimization scheduling (such as the MILP method), and in recent years, a two-stage collaborative optimization framework of "day-ahead + intraday" has emerged, but all lack a real-time correction layer at the second to minute level. Although existing MPC applications have time delay compensation, they have not yet formed a hierarchical closed-loop architecture with three time scales: day-ahead, intraday, and real-time. District cooling networks are long and have strong hydraulic time lags; the lack of real-time correction will lead to load response delays. District cooling involves various types of users, such as office buildings, commercial buildings, hotels, and residences, whose cooling load characteristics and cooling guarantee needs vary significantly. While existing research has proposed the concept of "load aggregator" and temperature-controlled load response model, it has not yet formed a user classification system and corresponding differentiated scheduling strategies for ice storage district cooling systems.

[0033] In view of this, this invention provides a method and related equipment for coordinated control of regional centralized cooling based on electric refrigeration and dynamic ice storage. First, multi-source data is collected, integrating meteorological, building operation, and user behavior data. A CNN-BiLSTM-Multi Head Self Attention hybrid model is used to accurately decompose and predict the cooling load components. Based on the prediction results and combined with time-of-use electricity pricing, a hierarchical optimization architecture with three time scales—day-ahead, intraday, and real-time—is adopted: the day-ahead layer generates a 24-hour day-ahead cooling source output plan with the minimum total cost; the intraday layer uses MPC rolling correction to address deviations; and the real-time layer combines feedforward compensation and PID fine-tuning to achieve second-level rapid response. Simultaneously, the dynamic ice storage device utilizes ultrasonic crystallization technology to reduce supercooling, efficiently generating flocculent ice crystals, and dynamically switches between storage and release modes based on load forecasts. Furthermore, distributed model predictive control achieves hydraulic balance in the pipe network and performs differentiated scheduling for users, ultimately participating in grid interaction as a virtual power plant resource to achieve global optimization of the reliability of the regional centralized cooling system.

[0034] The regional centralized cooling coordinated control method based on electric refrigeration and dynamic ice storage provided in this invention relates to the field of regional centralized cooling technology. This method can be applied to terminals, servers, or software running on either terminal or server. In some embodiments, the terminal can be a smartphone, tablet, laptop, desktop computer, smart speaker, smartwatch, or vehicle terminal, but is not limited to these. The server can be configured as an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms. The server can also be a node server in a blockchain network. The software can be an application implementing the regional centralized cooling coordinated control method based on electric refrigeration and dynamic ice storage, but is not limited to the above forms.

[0035] It should be noted that in various specific embodiments of the present invention, when processing data related to user identity or characteristics, such as user information, user behavior data, user historical data, and user location information, user permission or consent is obtained first. Furthermore, the collection, use, and processing of this data comply with relevant laws, regulations, and standards. In addition, when embodiments of the present invention require access to sensitive personal information of users, separate permission or consent from the user is obtained through pop-ups or redirection to a confirmation page. Only after obtaining the user's separate permission or consent is the necessary user-related data for the normal operation of the embodiments of the present invention acquired.

[0036] Figure 1 This is an optional flowchart of a regional centralized cooling coordinated control method based on electric refrigeration and dynamic ice storage provided in an embodiment of the present invention. Figure 1 The method may include, but is not limited to, steps S100 to S900: Step S100: Dynamically collect meteorological data, building operation data, user behavior data, pipeline operation status data, and equipment operation parameters; Step S200: Based on meteorological data, building operation data, and user behavior data, predict the hourly cooling load for the next 24 hours using a convolutional neural network, a bidirectional long short-term memory neural network, and a multi-head attention mechanism. Step S300: Obtain the day-ahead cooling source output plan for the next 24 hours based on the time-of-use electricity pricing mechanism; Step S400: Based on the pipeline operation status data and equipment operation parameters, dynamically obtain the setpoint sequence of cold source output for the next 4 hours; Step S500: Based on the pipeline network operation status data and the setpoint sequence of cold source output for the next 4 hours, the frequency of pipeline pumps and valve openings are finely adjusted using feedforward compensation and PID algorithm. Step S600: In response to the daytime cooling source output plan for the next 24 hours and the hourly cooling load for the next 24 hours, dynamically control the ice storage cooling. Step S700: In response to the hourly cooling load and pipeline operation status data for the next 24 hours, the hydraulic system of the pipeline network is dynamically balanced and controlled through a distributed model predictive control strategy. Step S800: Based on building operation data, classify users of various business types and dynamically adjust water supply temperature or cooling power according to the classification results. Step S900: In response to the hourly cooling load and equipment operating parameters for the next 24 hours, adjust the power consumption of the district cooling system.

[0037] In step S100 of some embodiments, multi-source data is collected in real time by a multi-dimensional cooling load prediction unit. This data may include, but is not limited to, meteorological data (including outdoor dry-bulb temperature, wet-bulb temperature, and solar radiation intensity), building operation data (including historical cooling load and air conditioning operation status of each user terminal), user behavior data (including heat map of personnel flow, equipment usage patterns, holiday arrangements, and business activity schedules), pipeline operation status data (including supply and return water temperature and instantaneous cooling load of each user terminal, flow rate and pressure of each branch, pressure difference between the primary and secondary sides of each heat exchange station, and indoor ambient temperature of users), equipment operation parameters (including real-time operating power and cooling capacity of electric chiller, ice storage tank level and ice slurry concentration), and regional economic data. The collected multi-source data is then written into the database in real time.

[0038] In step S200 of some embodiments, a time-series decomposition technique based on user behavior data is introduced to decompose the cold load sequence of each user into multiple load components. A hybrid model is used to predict each component, and the component prediction results are superimposed to output the hourly cold load prediction value of each user in the region for the next 24 hours and the total cold load prediction curve, providing uncertainty boundaries for optimized scheduling.

[0039] In some embodiments, step S200 may include, but is not limited to, steps S210 to S240: Step S210: Based on meteorological data, building operation data, and user behavior data, decompose the cooling load sequence of users in each business type to obtain load components; Step S220: Construct a hybrid model based on convolutional neural networks, bidirectional long short-term memory neural networks, and multi-head attention mechanisms; Step S230: Using a hybrid model, predict each load component to obtain the prediction results; Step S240: Overlay the prediction results to obtain the hourly cooling load for the next 24 hours.

[0040] In step S210 of some embodiments, the historical cooling load of each user terminal included in the building operation data is decomposed into three load components: "basic load" generated by the building envelope and fixed equipment, "meteorologically sensitive load" driven by outdoor temperature and humidity in meteorological data, and "behavioral load" driven by personnel movement and activity arrangements in user behavior data.

[0041] In step S220 of some embodiments, a CNN-BiLSTM-Multi Head Self Attention hybrid model can be constructed based on a convolutional neural network, a bidirectional long short-term memory neural network, and a multi-head attention mechanism.

[0042] In step S230 of some embodiments, a hybrid model is used to predict each load component to obtain the prediction result. For example, a convolutional neural network is constructed for each of the three load components to extract local features, a bidirectional long short-term memory neural network is constructed to capture bidirectional temporal dependencies, and a multi-head attention mechanism is fused to identify key time step patterns.

[0043] In step S240 of some embodiments, the prediction results of multiple load components are superimposed to generate hourly cooling load prediction values ​​and total cooling load prediction curves for the next 24 hours.

[0044] In step S300 of some embodiments, day-ahead optimization is performed in a three-time-scale hierarchical optimization architecture. Combined with time-of-use pricing, the objective function is to minimize the total operating cost of the regional centralized cooling system. The mixed-integer linear programming (MILP) method is used to solve for the day-ahead cooling source output plan for the next 24 hours. By combining time-of-use pricing with a cooling capacity distribution loss model for global optimization, the overall operating cost of the regional centralized cooling system is significantly reduced while meeting rigid cooling demand. Optionally, day-ahead optimization is performed at 00:00 daily on a 24-hour scale.

[0045] In some embodiments, step S300 may include, but is not limited to, steps S310 to S340: Step S310: Construct a cooling loss term based on the pipeline length, insulation thermal resistance, and supply and return water temperatures; Step S320: Construct an objective function based on the cooling loss term and the total operating cost of the regional centralized cooling system; Step S330: The output range of the electric refrigeration unit, the cold storage and release rate of the dynamic ice storage device, the upper and lower limits of the cold storage capacity of the ice storage tank, and the cold supply reliability constraints are used as limiting conditions. Step S340: Based on the objective function and constraints, use mixed integer linear programming to obtain the daytime cooling source output plan for the next 24 hours; The current cold source output plan includes the start / stop status and cooling power setting of each electric refrigeration unit, the ice storage capacity plan of the dynamic ice storage device, and the ice slurry release capacity plan.

[0046] In steps S310 to S320 of some embodiments, based on time-of-use electricity pricing, the objective function is to minimize the total operating cost of the regional centralized cooling system, and a cooling loss term Loss_network(t) is introduced into the objective function. This cooling loss term is related to the pipeline length, the thermal resistance of the insulation layer, and the temperature difference between the supply and return water, and can be calculated online through a pipeline thermal model.

[0047] In step S330 of some embodiments, the limiting conditions are the output range of the electric refrigeration unit, the cold storage and release rate of the dynamic ice storage device, the upper and lower limits of the cold storage capacity of the ice storage pool, and the cold supply reliability constraints (to ensure rigid user load requirements).

[0048] In step S340 of some embodiments, the mixed integer linear programming (MILP) method is used to solve the daytime cold source output plan for the next 24 hours based on the objective function and constraints. The solution outputs: the start-up and shutdown status and cooling power setpoint of each electric refrigeration unit (hourly resolution); the ice storage plan of the dynamic ice storage device (ice storage period and ice storage rate); and the ice slurry release plan (release period and release rate).

[0049] In step S400 of some embodiments, in the intraday rolling correction within the three-time-scale hierarchical optimization architecture, dynamic calibration of the day-ahead cooling source output plan can be achieved through model predictive control (MPC). Optionally, the intraday rolling correction is triggered every 4 hours, with an hourly time resolution. Based on the latest collected real-time meteorological data and load feedback, the day-ahead cooling source output plan is rolled over for optimization. This allows for adjustments to the output within a preset range to address load prediction deviations while tracking the day-ahead plan baseline. Ultimately, a sequence of cooling source output setpoints for the next 4 hours is output, providing baseline instructions to the real-time control layer. The load feedback includes pipeline operating status data and equipment operating parameters. Optionally, the load feedback may include, but is not limited to, the supply and return water temperatures and instantaneous cooling loads at each user terminal, the flow and pressure of each branch, the pressure difference between the primary and secondary sides of each heat exchange station, the user's indoor ambient temperature, the real-time operating power and cooling capacity of the electric chiller, and the ice storage tank level and ice storage capacity.

[0050] In some embodiments, step S400 may include, but is not limited to, steps S410 to S440: Step S410: Construct a state-space model based on the charging and discharging characteristics of the ice storage device; Step S420: Based on the pipeline operation status data and equipment operation parameters, the system status of the regional centralized cooling system at each time point in the next 4 hours is recursively predicted using a state space model. Step S430: Establish the mapping relationship between the preset physical constraints and the system state through the state-space model, and construct a quadratic programming problem based on the mapping relationship; Step S440: Obtain the solution to the quadratic programming problem and get the setpoint sequence of the cold source output for the next 4 hours.

[0051] In step S410 of some embodiments, a state-space model accurately characterizing the dynamic characteristics of the system is established, with the charging and discharging characteristics of the ice storage device as the core. Optionally, the system state variables of the state-space model include key parameters such as the ice storage capacity of the ice storage tank, the actual output of the electric refrigeration unit, and the supply and return water temperatures of the cooling system. The control input variables include the power adjustment command of the main unit, the frequency of the ice slurry delivery pump, and the valve opening. The output variable is the actual cooling capacity of each cold source node. The constructed model can accurately reflect the coupling relationship between the ice storage capacity, the cooling release rate, and the main unit output, providing mathematical model support for subsequent state prediction and optimization solutions. At the same time, the output of the model will serve as the input source of the feedforward compensator of the real-time correction layer, indirectly improving the second-level dynamic response accuracy.

[0052] In step S420 of some embodiments, a recursive prediction of the system state for the next 4 hours is performed based on the constructed state-space model. At the start of each MPC control cycle (every 4 hours), the current actual system state (e.g., current ice storage tank level, real-time operating power of the electric chiller, and measured supply and return water temperatures) is first read from the real-time database as the initial state. The system state for each moment in the next 4 hours is then recursively predicted hourly using the state transition equations of the state-space model. The prediction process must consider the hourly steady-state characteristics of the electric chiller (ignoring second-level dynamics, which are handled by the real-time layer) and incorporate the impact of pipeline transmission delays on state variables. The predicted state sequence (e.g., changes in ice storage volume and chiller output demand over the next 4 hours) will serve as input for subsequent constraint application and optimization, ensuring that the prediction results are consistent with the actual system dynamic characteristics.

[0053] In step S430 of some embodiments, a solvable quadratic programming problem is constructed by mapping physical constraints to the predicted system state. For example, physical constraints such as upper and lower limits of ice storage capacity (e.g., maximum / minimum cold storage capacity of the ice storage tank), limits on the rate of change of main unit output (e.g., maximum adjustment range of main unit power per minute), and supply and return water temperature ranges (e.g., supply water temperature not lower than 5°C, return water temperature not higher than 15°C) are mapped to the predicted system state for the next 4 hours through the output equation of the state-space model, forming a set of inequality constraints. Simultaneously, with the goal of tracking the daily plan, the objective function is defined as minimizing the sum of squared deviations between the predicted output and the daily planned output, combined with the system energy consumption cost term to construct a quadratic objective function. This constraint mapping step ensures that the optimization result meets the equipment's safe operating boundary, avoids the predicted state from exceeding the physical feasible region, and provides a rigorous mathematical optimization framework for subsequent solutions.

[0054] In step S440 of some embodiments, a sequence of setpoint values ​​for the cold source output over the next 4 hours is obtained by solving a quadratic programming problem, and the first instruction is executed. For example, based on the constructed quadratic programming problem, a numerical optimization algorithm (such as the interior-point method) is used to solve for the optimal control sequence over the next 4 hours. The sequence includes the setpoint power of the electric chiller for each hour, the frequency of the ice slurry transfer pump, and mode switching instructions (such as "chiller-only cooling" or "ice slurry release + chiller combined cooling"). According to the MPC rolling optimization mechanism, only the control instruction for the first hour of the sequence (hours 0-1) is sent to the field equipment via a distributed control system (DCS) or programmable logic controller (PLC) to adjust the chiller guide vane opening, inverter frequency, and electric valve opening, so that the actual output reaches the setpoint value. After execution, the next sampling cycle begins (4 hours later). The initial state of the state space model is updated based on the new measured data. The prediction-optimization-execution process is repeated, which means returning to the execution step of recursively predicting the system state of the regional centralized cooling system at each time in the next 4 hours based on the pipeline operation status data and equipment operation parameters, thus forming a closed-loop rolling optimization.

[0055] In step S500 of some embodiments, in the real-time correction of the three-time-scale hierarchical optimization architecture, the system's overall response time to disturbances is ensured to be less than or equal to 30 seconds by addressing short-term random fluctuations in load (such as sudden changes in personnel density or equipment start-up and shutdown). This step runs continuously with a sampling period of 1 second. Based on the setpoint sequence of the cooling source output for the next 4 hours output by the intraday rolling correction layer, combined with real-time collected pipeline network operation status data, a composite control strategy combining improved PID and feedforward compensation is used to fine-tune the frequency of pipeline pumps and valve openings. This compensates for the dynamic response lag of the intraday hourly optimization. Through the synergy of predictive adjustment and feedback correction, the impact of hydraulic imbalance on cooling quality is eliminated, achieving precise delivery of cooling capacity and rapid stabilization of terminal temperatures.

[0056] In some embodiments, step S500 may include, but is not limited to, steps S510 to S530: Step S510: Perform linear interpolation on the setpoint sequence of cold source output for the next 4 hours to obtain the predicted value of cold source output change for the next 15 minutes as the feedforward compensation value. Step S520: Collect the supply and return water temperatures from the pipeline operation status data once per second, obtain the deviation between the supply and return water temperatures and the set values, and output the feedback correction value through the PID algorithm. Step S530: The feedforward compensation value and the feedback correction value are superimposed to generate an execution command to adjust the frequency of the pipeline pumps and the valve opening.

[0057] In step S510 of some embodiments, the hourly cold source output setpoint sequence of the intraday layer is converted into a high-frequency prediction signal available to the real-time layer. For example, every 5 minutes, the cold source output setpoint sequence for the next 4 hours (such as main unit power, ice slurry flow rate, mode switching command) is obtained from the intraday MPC. For the output change in the first hour (k=0 to k=1) of the sequence, a starting value P(0) and an ending value P(1) are taken. It is assumed that the output changes linearly within this hour (since the dynamic response of the electric chiller is close to first-order inertia, a linear approximation within 15 minutes is sufficiently accurate). The predicted value for every 5 minutes within the next 15 minutes is calculated using linear interpolation. Further, the predicted output change for the next 15 minutes is calculated as P(15 minutes) - P(0), which represents the amount by which the cold source output needs to increase or decrease within the next 15 minutes according to the MPC plan. This predicted value serves as a feedforward compensation input, enabling the system to adjust the pump frequency in advance and avoid passive responses caused by sudden load changes.

[0058] In step S520 of some embodiments, the control output is dynamically adjusted by monitoring the cooling effect at the terminal in real time. With a sampling period of 1 second, temperature sensors located in each user's heat exchanger room collect the supply and return water temperatures in real time, calculate the difference between them to obtain the terminal supply and return water temperature difference, and compare this temperature difference with a set value (e.g., 5°C) to obtain the deviation. An improved PID algorithm (e.g., incorporating derivative-first or integral separation strategies) is used to process this deviation, and a feedback correction value is output. The function of the PID controller is to eliminate the residual deviation after feedforward compensation, ensuring that the terminal supply and return water temperature difference remains stable within the set range, avoiding temperature fluctuations caused by feedforward prediction errors or external disturbances. Its 1-second sampling period ensures the real-time performance and precision of the control.

[0059] In step S530 of some embodiments, the combined result of feedforward and feedback is converted into adjustment actions of field equipment. The feedforward compensation value and the feedback correction value are superimposed to generate the final control command (e.g., "increase the frequency of the ice slurry transfer pump by 5Hz, from 35Hz to 40Hz"). The command is then sent to the field actuators (frequency converters, electric valves) via DCS or PLC to adjust the frequency of the pipeline pumps and the valve opening, thereby achieving dynamic balance of cooling capacity. This step updates the feedforward compensation value every 5 minutes and continuously outputs the PID feedback correction value at a frequency of 1 second, forming a dual closed-loop control of feedforward prediction and feedback correction. This ensures that the system's response time to load disturbances is less than or equal to 30 seconds, ultimately achieving hydraulic balance of the pipeline network and stability of the terminal cooling quality.

[0060] In step S600 of some embodiments, by closely linking the day-ahead cold source output plan with the hourly cooling load forecast for the next 24 hours, and based on the time-of-use electricity pricing mechanism and ultrasonic crystallization technology, efficient cold storage is ensured during off-peak hours and priority is given to cold use during peak hours. At the same time, load fluctuations are handled flexibly through multiple modes, thereby balancing cold storage costs, cooling reliability and system energy consumption, avoiding the common problems of excessive ice storage waste or insufficient cooling capacity shutdown in traditional ice storage systems, and maximizing the stability of the district cooling system.

[0061] In some embodiments, step S600 may include, but is not limited to, steps S610 to S640: Step S610: During periods of low electricity prices, in response to the ice storage plan in the daytime cold source output plan for the next 24 hours, the electric refrigeration unit is switched to ice-making mode, and the ultrasonic crystallization unit is activated to reduce the supercooling of ice formation and generate flocculent ice crystals to be stored in the ice slurry storage tank. Step S620: During peak electricity price periods, in response to the hourly cooling load of the next 24 hours, prioritize the release of ice slurry from the ice slurry storage tank. Step S630: When the hourly cooling load for the next 24 hours exceeds the ice slurry release capacity, automatically switch to the combined cooling mode of electric refrigeration unit and ice slurry release. In step S640, during the ice slurry release process, the frequency of the ice slurry delivery pump and the valve opening are dynamically adjusted based on the supply and return water temperatures in the pipeline operation status data.

[0062] In step S610 of some embodiments, an efficient ice-making process is initiated during off-peak electricity hours (valley electricity hours). Specifically, the dynamic ice storage intelligent scheduling unit responds to the ice storage plan in the day-ahead cooling source output plan. First, it controls the dual-mode electric refrigeration unit to seamlessly switch from conventional refrigeration mode to ice-making mode. Simultaneously, it activates the ultrasonic crystallization unit, which uses high-frequency vibration to disrupt the surface tension of the supercooled water, significantly reducing the ice supercooling (from the conventional -8°C to -10°C to -2°C to -4°C). This promotes direct heat exchange between the refrigerant and water within the heat exchange tubes, generating flocculent ice crystals with high porosity. These flocculent ice crystals, due to their excellent fluidity, naturally settle and are stored in the ice slurry storage pool. The total ice storage volume is strictly determined based on the predicted total cooling load for the next day, adhering to the principle of on-demand cooling storage, ensuring peak cooling demand for the next day while avoiding ineffective energy consumption.

[0063] In step S620 of some embodiments, stored cooling capacity is prioritized during peak electricity price periods (peak electricity periods) to reduce operating costs. This step responds to hourly cooling load forecasts for the next 24 hours. When it is detected that the current period is a peak electricity period and there is cooling load demand, the release valve of the ice slurry storage tank is automatically opened. Utilizing the characteristic of direct heat exchange between the ice slurry and the chilled water return water, the cooling capacity is released rapidly. Compared to traditional coil-type ice, flocculent ice crystals have a huge specific surface area, enabling instantaneous heat exchange with the return water. The load response speed is extremely fast, effectively smoothing out the instantaneous surge in peak load. The scheduling logic strictly follows the principle of prioritizing cooling release during peak hours and releasing cooling as needed during normal times. Only when the real-time cooling load is extremely low or the cooling capacity is insufficient will the electric chiller be activated as an auxiliary unit, maximizing the utilization of the cooling capacity stored in off-peak electricity.

[0064] In step S630 of some embodiments, combined cooling is activated when a single cold source cannot meet the demand. This step monitors the hourly cooling load forecast for the next 24 hours and the real-time cooling capacity of the current ice slurry storage tank: when the predicted cooling load exceeds the maximum cooling capacity threshold of the ice slurry, the system automatically triggers a combined cooling mode of electric chiller and ice slurry cooling. At this time, the dual-mode chiller resumes cooling mode and shares the load with the ice slurry; when the cooling load demand is low (such as at night or during transitional seasons), the system automatically switches to a single mode of cooling only by electric chiller or only by ice slurry cooling, avoiding energy waste. This dynamic switching strategy based on load forecasting ensures that the cold source always operates in the high-efficiency range, improving the overall system energy efficiency ratio (COP).

[0065] In step S640 of some embodiments, the distribution efficiency is dynamically optimized during the ice slurry release process. This step uses the real-time chilled water return temperature of each user terminal, collected from the pipeline network operation status data, as a feedback signal for the release intensity: when a sudden increase in the load at the user terminal causes the return water temperature to rise, the frequency of the ice slurry delivery pump is automatically increased and the corresponding valves are opened wider to increase the ice slurry flow rate and enhance heat exchange; when the load decreases and the return water temperature drops, the system is adjusted in the opposite direction to reduce cooling waste. Through this closed-loop regulation based on return water temperature, direct and efficient heat exchange between the ice slurry and the return water is achieved, which not only ensures the stability of the terminal cooling temperature but also avoids the ineffective power consumption of the ice slurry delivery pump, further refining the energy-saving benefits of the dynamic ice storage system.

[0066] In step S700 of some embodiments, to address the strong hydraulic coupling and large transmission time delay in long-distance pipeline networks, a distributed model predictive control (DMPC) strategy is used to achieve dynamic hydraulic balance in the network. For example, in response to hourly cooling load forecasts for the next 24 hours and real-time collected pipeline operation status data, a master-slave collaborative control architecture is constructed. The model predictive control (MPC) on the cooling source side acts as the master controller, responsible for deciding the total cooling capacity setpoint based on global load demand; the DMPC on the pipeline side acts as the slave controller, not independently deciding the cooling capacity, but focusing on finely decomposing and executing the total volume commands issued by the master controller in the spatiotemporal dimensions. This architecture effectively solves the problem of overheating for near-end users and undercooling for far-end users caused by hydraulic imbalance in long-distance pipeline networks, significantly reducing ineffective energy consumption in the transmission and distribution system while ensuring the quality of end-point cooling.

[0067] In some embodiments, step S700 may include, but is not limited to, steps S710 to S740: Step S710: Use the hourly cooling load of the next 24 hours as feedforward input to adjust the opening of the regulating valves and the frequency of the circulating pumps in each pipe network branch. Step S720: Based on the flow and pressure of each branch and the pressure difference of each heat exchange station in the pipeline network operation status data, the opening degree of the regulating valve of each pipeline branch is coordinated and adjusted through a distributed model predictive control strategy.

[0068] In step S710 of some embodiments, high-precision load forecasting is used to intervene in the pipeline network status in advance. For example, the hourly cooling load curve for the next 24 hours output by the multi-dimensional cooling load forecasting unit is used as a feedforward input signal. Based on the trend of load changes (such as load increases before the morning peak), the opening of regulating valves and the frequency of circulating pumps in each pipeline branch are adjusted several minutes to tens of minutes in advance. For instance, when it is predicted that a peak in pedestrian traffic and a surge in cooling load demand are about to occur in a certain area, the system opens the electric regulating valve of that branch in advance and increases the speed of the variable frequency pump, so that the hydraulic and thermal conditions of the pipeline network respond in advance, thereby physically eliminating the transmission lag caused by excessively long pipelines and limited water flow velocity. This prediction-based proactive adjustment ensures that cooling capacity is accurately delivered to the required location at the required time.

[0069] In step S720 of some embodiments, distributed coordinated regulation is performed based on real-time status feedback. For example, the flow and pressure of each branch and the pressure difference between the primary and secondary sides of each heat exchange station are continuously monitored in the pipeline network operation status data. The supply and return water temperature difference at each user terminal (reflecting actual cooling demand) is used as the target control variable to drive a distributed model predictive control strategy. Unlike traditional centralized control, DMPC allows local controllers of each branch to independently calculate and optimize valve regulation based on local real-time pressure difference and flow feedback after receiving the total control command from the main controller. Simultaneously, coordinated compensation is achieved through information exchange between adjacent controllers (such as pressure fluctuations in adjacent branches). This distributed coordinated regulation method enables on-demand allocation of cooling capacity and prevents energy waste.

[0070] In step S800 of some embodiments, users are classified and managed in a refined manner based on building operation data (such as historical cooling load and air conditioning operation status of each user terminal) to achieve differentiated scheduling of cooling resources. First, users are divided into three categories: rigid load users (such as data centers and hospitals that require continuous and stable cooling), adjustable load users (such as office buildings and shopping malls that have a certain degree of load flexibility), and flexible load users (such as residential buildings and hotels whose load demand can be flexibly adjusted during certain periods). During peak grid load periods, priority is given to reducing the cooling supply to adjustable load users and flexible load users by appropriately increasing the water supply temperature or reducing the cooling power (e.g., increasing the water supply temperature of office buildings from 7°C to 9°C, or reducing the number of air conditioning units operating in shopping malls) to ensure that the cooling quality of rigid load users is not affected. Simultaneously, based on users' willingness and ability to participate in demand response, a user response incentive mechanism (such as electricity discounts and priority service guarantees) is established to guide users to actively cooperate with system scheduling.

[0071] In step S900 of some embodiments, the regional centralized cooling system is connected to the virtual power plant platform as an adjustable load aggregation unit to achieve coordinated interaction with the power grid. Based on the hourly cooling load forecast for the next 24 hours and equipment operating parameters (such as the real-time operating power of the electric chiller, the ice storage tank level, and the ice slurry concentration), day-ahead and intraday active power responses are executed: In the day-ahead active power response phase, a day-ahead reporting strategy (such as reporting the load reduction amount during the peak period of the power grid the next day) is formed according to the day-ahead load forecast of the cooling station and the day-ahead response requirements of the virtual power plant management center (such as the peak shaving requirements of the power grid), and the strategy is sent to the cooling station. The power consumption of the cooling system is dynamically controlled by adjusting the power of the electric chiller, the ice slurry release rate, etc. In the intraday active power response phase, the power consumption is dynamically adjusted according to the intraday load forecast of the cooling station and the intraday response requirements of the management center (such as the real-time power grid balance requirements) (such as rapidly reducing the cooling power of non-rigid load areas when there is a sudden power grid gap), ensuring the reliability of cooling supply.

[0072] This invention also provides a regional centralized cooling coordinated control device based on electric refrigeration and dynamic ice storage, which can realize the above-mentioned regional centralized cooling coordinated control method based on electric refrigeration and dynamic ice storage. The device includes: The multi-source data acquisition module is used to dynamically collect meteorological data, building operation data, user behavior data, pipeline operation status data, and equipment operation parameters; The cooling load forecasting module is used to predict the hourly cooling load for the next 24 hours based on meteorological data, building operation data, and user behavior data, using convolutional neural networks, bidirectional long short-term memory neural networks, and multi-head attention mechanisms. The day-ahead optimization module is used to obtain the day-ahead cooling source output plan for the next 24 hours based on the time-of-use electricity pricing mechanism; The intraday rolling correction module is used to dynamically obtain the sequence of cold source output setpoints for the next 4 hours based on pipeline operation status data and equipment operating parameters. The real-time correction module is used to fine-tune the frequency of water pumps and valve openings in the pipeline network based on the pipeline network operation status data and the set value sequence of cold source output for the next 4 hours, through feedforward compensation and PID algorithm. The dynamic ice storage intelligent scheduling module is used to dynamically control ice storage in response to the day-ahead cooling source output plan and hourly cooling load for the next 24 hours. The pipeline hydraulic balance control module is used to dynamically balance the pipeline hydraulics in response to hourly cooling load and pipeline operation status data for the next 24 hours through a distributed model predictive control strategy. The user load response and differentiated scheduling module is used to classify users of various business types based on building operation data, and dynamically adjust the water supply temperature or cooling power according to the classification results. The virtual power plant collaborative control module is used to adjust the power consumption of the district cooling system in response to the hourly cooling load and equipment operating parameters for the next 24 hours.

[0073] It is understood that the content of the above method embodiments is applicable to the present device embodiments. The specific functions implemented by the present device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0074] This invention also provides an electronic device, which includes a processor and a memory. The memory stores a computer program, and the processor executes the computer program to implement the above-described method. This electronic device can be any smart terminal, including a tablet computer, an in-vehicle computer, or similar device.

[0075] It is understood that the content of the above method embodiments is applicable to this device embodiment. The specific functions implemented by this device embodiment are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0076] refer to Figure 2 , Figure 2 The hardware structure of an electronic device according to another embodiment is illustrated. The electronic device includes: The processor 1001 can be implemented using a general-purpose central processing unit (CPU), microprocessor, application specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of the present invention. The memory 1002 can be implemented as a read-only memory (ROM), static storage device, dynamic storage device, or random access memory (RAM). The memory 1002 can store the operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 1002 and is called and executed by the processor 1001. Input / output interface 1003 is used to implement information input and output; The communication interface 1004 is used to enable communication and interaction between this device and other devices. Communication can be achieved through wired means (such as USB, network cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.). Bus 1005 transmits information between various components of the device (e.g., processor 1001, memory 1002, input / output interface 1003, and communication interface 1004); The processor 1001, memory 1002, input / output interface 1003 and communication interface 1004 are connected to each other within the device via bus 1005.

[0077] This invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method.

[0078] It is understood that the content of the above method embodiments is applicable to this storage medium embodiment. The specific functions implemented in this storage medium embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved in the above method embodiments.

[0079] This invention also provides a computer program product or computer program that includes computer instructions stored in a computer-readable storage medium. A processor of a computer device can read the computer instructions from the computer-readable storage medium and execute the computer instructions to cause the computer device to perform the aforementioned method.

[0080] In summary, the regional centralized cooling coordinated control method and related equipment based on electric refrigeration and dynamic ice storage of the present invention have the following advantages: 1. This invention integrates multi-source heterogeneous data such as meteorological, building operation and user behavior, and innovatively decomposes the cooling load into three components: "basic, meteorological sensitive and behavioral" for independent modeling and prediction. This effectively removes complex interference factors and solves the problem that traditional methods cannot cope with random load fluctuations, providing a high-confidence data foundation for subsequent optimization scheduling.

[0081] 2. The embodiments of the present invention combine a three-level closed-loop optimization architecture of day-day, intraday, and real-time. It not only reduces operating costs by utilizing the time-of-use electricity pricing mechanism through cold storage and peak shaving, but also compresses the system's response time to load disturbances to within 30 seconds through a second-level real-time correction layer (feedforward + PID), thus taking into account both macroeconomic efficiency and the stability of micro-level cooling quality.

[0082] 3. The embodiments of the present invention reduce the supercooling of ice formation by introducing ultrasonic crystallization technology, which solves the problems of high energy consumption and easy freezing in traditional ice making; at the same time, based on load prediction, the single cooling supply / single cooling release / combined cooling supply modes are dynamically switched to avoid excessive activation of the cold source or waste of cold energy, and realize on-demand cold storage and precise cold release.

[0083] 4. The embodiments of the present invention adopt a distributed model predictive control (DMPC) strategy based on predictive feedforward, which uses load forecasting to adjust the pipeline valves and pump frequencies in advance, eliminating the physical lag of long-distance transmission and distribution; combined with a master-slave control architecture, it realizes the on-demand allocation of cooling capacity of each branch without changing the total cooling capacity, which significantly reduces transmission and distribution energy consumption.

[0084] 5. By aggregating and connecting the district cooling system as an adjustable resource to a virtual power plant, and participating in day-ahead and intraday active power response, this invention can not only proactively reduce load during peak grid periods to obtain subsidy benefits, but also flexibly adjust the cooling priority (rigid / adjustable / flexible) of different business users, and prioritize the cooling security of critical users (such as hospitals and data centers) under extreme operating conditions.

[0085] In some alternative embodiments, the functions / operations mentioned in the block diagrams may not occur in the order shown in the operation diagrams. For example, depending on the functions / operations involved, two consecutively shown blocks may actually be executed substantially simultaneously, or the blocks may sometimes be executed in reverse order. Furthermore, the embodiments presented and described in the flowcharts of this invention are provided by way of example to provide a more comprehensive understanding of the technology. The disclosed methods are not limited to the operations and logic flows presented herein. Alternative embodiments are contemplated in which the order of various operations is altered and sub-operations described as part of a larger operation are executed independently.

[0086] Furthermore, although the invention has been described in the context of functional modules, it should be understood that, unless otherwise stated, one or more of the described functions and / or features may be integrated into a single physical device and / or software module, or one or more functions and / or features may be implemented in a separate physical device or software module. It is also understood that a detailed discussion of the actual implementation of each module is unnecessary for understanding the invention. Rather, given the properties, functions, and internal relationships of the various functional modules in the apparatus disclosed herein, the actual implementation of the module will be understood within the scope of conventional skill of an engineer. Therefore, those skilled in the art can implement the invention as set forth in the claims using ordinary techniques without excessive experimentation. It is also understood that the specific concepts disclosed are merely illustrative and not intended to limit the scope of the invention, which is determined by the full scope of the appended claims and their equivalents.

[0087] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory, random access memory, magnetic disks, or optical disks.

[0088] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-including system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0089] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0090] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0091] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

[0092] The above is a detailed description of the preferred embodiments of the present invention. However, the present invention is not limited to the embodiments described. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the present invention. All such equivalent modifications or substitutions are included within the scope defined by the claims of the present invention.

Claims

1. A method for coordinated control of regional centralized cooling based on electric refrigeration and dynamic ice storage, characterized in that, Includes the following steps: Dynamically collect meteorological data, building operation data, user behavior data, pipeline operation status data, and equipment operation parameters; Based on the meteorological data, the building operation data, and the user behavior data, the hourly cooling load for the next 24 hours is predicted using a convolutional neural network, a bidirectional long short-term memory neural network, and a multi-head attention mechanism. Based on the time-of-use electricity pricing mechanism, obtain the day-ahead cooling source output plan for the next 24 hours; Based on the pipeline operation status data and equipment operation parameters, dynamically obtain the cold source output setpoint sequence for the next 4 hours; Based on the pipeline network operation status data and the setpoint sequence of the cold source output for the next 4 hours, the frequency of the pipeline pumps and the valve opening are finely adjusted through feedforward compensation and PID algorithm. The ice storage is dynamically controlled in response to the daytime cooling source output plan for the next 24 hours and the hourly cooling load for the next 24 hours. In response to the hourly cooling load for the next 24 hours and the pipeline network operation status data, a distributed model predictive control strategy is used to dynamically balance the hydraulic system of the pipeline network. Based on the building operation data, users of various business types are classified, and the water supply temperature or cooling power is dynamically adjusted according to the classification results. The power consumption of the district cooling system is adjusted in response to the hourly cooling load for the next 24 hours and the equipment operating parameters.

2. The method according to claim 1, characterized in that, The method of predicting hourly cooling load for the next 24 hours based on meteorological data, building operation data, and user behavior data, using a convolutional neural network, a bidirectional long short-term memory neural network, and a multi-head attention mechanism, includes the following steps: Based on the meteorological data, the building operation data, and the user behavior data, the cold load sequence of each of the aforementioned business types is decomposed to obtain load components; A hybrid model is constructed based on the convolutional neural network, the bidirectional long short-term memory neural network, and the multi-head attention mechanism; The hybrid model is used to predict each load component, and the prediction results are obtained. The prediction results are superimposed to obtain the hourly cooling load for the next 24 hours.

3. The method according to claim 1, characterized in that, The process of obtaining the day-ahead cooling source output plan for the next 24 hours based on the time-of-use electricity pricing mechanism includes the following steps: Based on the pipeline length, thermal resistance of the insulation layer, and supply and return water temperatures, a cooling loss term is constructed. Based on the cooling loss term and the total operating cost of the regional centralized cooling system, construct the objective function; The output range of the electric refrigeration unit, the cold storage and release rate of the dynamic ice storage device, the upper and lower limits of the cold storage capacity of the ice storage tank, and the constraints on the reliability of the cooling supply are used as limiting conditions. Based on the objective function and the constraints, a mixed-integer linear programming approach is used to obtain the daytime cooling source output plan for the next 24 hours. The daytime cold source output plan includes the start / stop status and cooling power setting of each electric refrigeration unit, the ice storage capacity plan of the dynamic ice storage device, and the ice slurry release capacity plan.

4. The method according to claim 1, characterized in that, The step of dynamically obtaining the setpoint sequence of cooling source output for the next 4 hours based on the pipeline network operation status data and equipment operation parameters includes the following steps: Based on the charging and discharging characteristics of the ice storage device, a state-space model is constructed; Based on the pipeline operation status data and equipment operation parameters, the system status of the regional centralized cooling system at each time point in the next 4 hours is recursively predicted through the state space model. Using the state-space model, a mapping relationship between preset physical constraints and the system state is established, and a quadratic programming problem is constructed based on the mapping relationship. The solution to the quadratic programming problem is obtained, and the setpoint sequence of the cold source output for the next 4 hours is obtained.

5. The method according to claim 1, characterized in that, The step of fine-tuning the frequency of water pumps and valve openings in the pipeline network based on the pipeline network operation status data and the setpoint sequence of cold source output for the next 4 hours, using feedforward compensation and PID algorithm, includes the following steps: Linear interpolation is performed on the setpoint sequence of cold source output for the next 4 hours to obtain the predicted value of cold source output change for the next 15 minutes as the feedforward compensation value. The supply and return water temperatures in the pipeline network operation status data are collected once per second. The deviation between the supply and return water temperatures and the set values ​​is obtained, and a feedback correction value is output through a PID algorithm. The feedforward compensation value and the feedback correction value are superimposed to generate an execution command to adjust the frequency of the pipeline pumps and the valve opening.

6. The method according to claim 1, characterized in that, The dynamic control of ice storage in response to the daytime cooling source output plan for the next 24 hours and the hourly cooling load for the next 24 hours includes the following steps: During periods of low electricity prices, in response to the ice storage plan in the daytime cold source output plan for the next 24 hours, the electric refrigeration unit is switched to ice-making mode, and the ultrasonic crystallization unit is activated to reduce the supercooling of ice formation, generating flocculent ice crystals and storing them in the ice slurry storage tank. During peak electricity price periods, in response to the hourly cooling load of the next 24 hours, priority is given to controlling the ice slurry storage tank to release ice slurry for cooling. When the hourly cooling load for the next 24 hours exceeds the ice slurry release capacity, it will automatically switch to a combined cooling mode of electric refrigeration unit and ice slurry release. During the ice slurry release process, the frequency of the ice slurry delivery pump and the valve opening are dynamically adjusted based on the supply and return water temperatures in the pipeline operation status data.

7. The method according to claim 1, characterized in that, The system, responding to the hourly cooling load data for the next 24 hours and the pipeline network operating status data, uses a distributed model predictive control strategy to dynamically balance the hydraulic system of the pipeline network, including the following steps: The hourly cooling load for the next 24 hours is used as a feedforward input to adjust the opening of the regulating valves and the frequency of the circulating pumps in each branch of the pipe network. Based on the flow and pressure of each branch and the pressure difference of each heat exchange station in the pipeline network operation status data, the opening of the regulating valves of each branch of the pipeline network is coordinated and adjusted through a distributed model predictive control strategy.

8. A regional centralized cooling coordinated control device based on electric refrigeration and dynamic ice storage, characterized in that, include: The multi-source data acquisition module is used to dynamically collect meteorological data, building operation data, user behavior data, pipeline operation status data, and equipment operation parameters; The cooling load prediction module is used to predict the hourly cooling load for the next 24 hours based on the meteorological data, the building operation data, and the user behavior data, using a convolutional neural network, a bidirectional long short-term memory neural network, and a multi-head attention mechanism. The day-ahead optimization module is used to obtain the day-ahead cooling source output plan for the next 24 hours based on the time-of-use electricity pricing mechanism; The intraday rolling correction module is used to dynamically obtain the sequence of cold source output setpoints for the next 4 hours based on the pipeline operation status data and equipment operation parameters. The real-time correction module is used to fine-tune the frequency of the water pumps and the valve opening based on the pipeline network operation status data and the set value sequence of the cold source output for the next 4 hours, through feedforward compensation and PID algorithm. The dynamic ice storage intelligent scheduling module is used to dynamically control ice storage in response to the daytime cold source output plan for the next 24 hours and the hourly cooling load for the next 24 hours. The pipeline hydraulic balance control module is used to dynamically balance the pipeline hydraulics in response to the hourly cooling load and pipeline operation status data for the next 24 hours, through a distributed model predictive control strategy. The user load response and differentiated scheduling module is used to classify users of various business types according to the building operation data, and dynamically adjust the water supply temperature or cooling power according to the classification results. The virtual power plant collaborative control module is used to adjust the power consumption of the district cooling system in response to the hourly cooling load and the equipment operating parameters for the next 24 hours.

9. An electronic device, characterized in that, Including the processor and memory; The memory is used to store programs; The processor executes the program to implement the method as described in any one of claims 1 to 7.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 7.

Images