Intelligent management and control system and method for refrigerating machine room based on dynamic correction

Abstract

The invention discloses a refrigerating machine room intelligent management and control system and method based on dynamic correction, and relates to the technical field of intelligent management and control. The system comprises a data acquisition module, a model management and control module, a scheme optimization module and an instruction execution module. The method comprises the steps of collecting and verifying full-link equipment operation data of a refrigerating machine room, and generating an initial adjustment scheme; deducing a linkage failure path caused by an adjustment action in the management and control model and quantifying the probability, rejecting over-limit adjustment content in combination with a dynamic failure probability threshold, and obtaining a target adjustment scheme through collaborative verification iterative optimization; and based on the failure rule correction model parameters, converting the target scheme into a time sequence regulation and control instruction to be issued and executed, and forming closed-loop management and control through execution effect verification. According to the method, the failure risk can be pre-judged and adjusted in advance, equipment regulation and control conflicts and cascading failures are avoided, the adaptive capacity of the model is improved, the operation energy efficiency and stability of the refrigerating machine room are improved, and the operation requirements of the efficient refrigerating machine room are met.

Description

Technical Field

[0001] This invention relates to the field of intelligent control technology, specifically to an intelligent control system and method for refrigeration rooms based on dynamic correction. Background Technology

[0002] The building's refrigeration room is the core of the central air conditioning system's energy consumption. To achieve energy-saving operation, existing technologies mostly use the Internet of Things, data mining, and other means to collect the room's operating data, calculate the optimal adjustment scheme, and then output positive equipment control commands to achieve automatic adjustment and dynamic correction of the refrigeration room.

[0003] However, this positive adaptation logic has significant technical flaws and is difficult to meet the actual operational requirements of high-efficiency cooling room: the existing control model lacks the ability to predict potential failure scenarios of the proposed output adaptation scheme, and only verifies the effect after the command is executed. This can easily lead to negative effects such as decreased equipment efficiency and reduced system energy efficiency. In unattended high-efficiency cooling room scenarios, it is even more likely to trigger a chain of equipment failures. At the same time, the existing technology does not analyze the chain reaction after the adaptation command is executed. Adjustment actions targeting only local equipment are prone to triggering system-level operational failures, making it impossible to achieve the goal of optimizing the cooling room from local to overall. In addition, the existing control does not incorporate the patterns of adaptation failures into the dynamic correction basis of the control model, resulting in the recurrence of similar failure problems. The adaptive capability of the control model is weak, and the trial and error cost of equipment adaptation remains high.

[0004] The aforementioned problems directly result in insufficient effectiveness and stability of intelligent management and control of refrigeration rooms, making it difficult to meet the core requirements of achieving energy efficiency standards and uninterrupted reliable operation of high-efficiency refrigeration rooms throughout the year. Therefore, it is urgent to optimize the existing management and control logic and propose an intelligent management and control solution for refrigeration rooms that can predict adjustment failures in advance and avoid negative adjustment actions. Summary of the Invention

[0005] The purpose of this invention is to provide an intelligent management and control system and method for refrigeration rooms based on dynamic correction, so as to solve the problems mentioned in the background art.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: A method for intelligent management and control of a chiller room based on dynamic correction includes the following steps: S1. Collect real-time operating data of all equipment in the chiller room, input the real-time operating data into the chiller room dynamic correction and control model for operating status analysis, and generate an initial adjustment plan for the equipment in the chiller room based on the analysis results; S2. In the dynamic correction and control model of the cooling room, the existing operating status of the room is restored by combining real-time operating data. For each adjustment action in the initial adjustment plan, the failure path caused by the chain reaction between the devices in the entire link of the room after its execution is deduced one by one, and the probability of occurrence of each failure path is quantified. S3. Preset a failure probability threshold, remove the adjustment content corresponding to failure paths whose occurrence probability exceeds the failure probability threshold, iteratively optimize the initial adjustment plan to obtain the target adjustment plan; use the failure paths of the entire link and the corresponding probability rules as the basis for dynamic correction, update the parameters of the dynamic correction and control model of the chiller room, and complete the dynamic correction of the model. S4. Transform the target adjustment plan into standardized equipment control commands, output the equipment control commands to the actuators in the refrigeration room, drive the corresponding equipment in the room to perform control actions, and realize intelligent management and control of the refrigeration room.

[0007] Furthermore, S1 includes the following: Real-time operating data of all equipment in the refrigeration room is collected. The equipment in the entire chain includes chillers, cooling towers, cooling pumps, and chilled pumps. The real-time operating data includes the operating power, medium temperature, medium flow rate, and number of devices in operation for each device. At the same time, the total cooling capacity of the computer room system is collected. The validity of the collected real-time operation data is verified by using the parameter fluctuation threshold verification method commonly used in the field of refrigeration room. Abnormal data and missing data that exceed the normal fluctuation range are removed to obtain a standardized data set of refrigeration room operation data, which is then input into the dynamic correction and control model of refrigeration room. The dynamic correction and control model for the chiller room analyzes the operating status based on the data set of the chiller room operation, and calculates the real-time comprehensive energy efficiency ratio of the chiller room system and the real-time operating load rate of each device in sequence. The real-time operating parameters of the chiller, cooling pump and chilled pump are the medium flow rate, and the real-time operating parameters of the cooling tower are the water treatment volume. The dynamic calibration and control model for the chiller room aims to maximize the real-time comprehensive energy efficiency ratio of the chiller room system. It combines the real-time operating load rate of each device and the device linkage operation rules to match the cooling capacity requirements of the chiller room and generate an initial adjustment plan for the joint control of single or multiple devices in the chiller room. The adjustment actions of the initial adjustment plan include adjusting the device frequency, increasing or decreasing the number of devices in operation, and fine-tuning the device operating parameters. The dynamic calibration and control model for the chiller room is a preset simulation analysis model adapted to the operation and control of the entire chain of devices in the chiller room.

[0008] Furthermore, S2 includes the following: In the dynamic correction and control model of the chiller room, the real-time coupled operation status of the entire link equipment in the chiller room is restored by combining the standardized data set of the chiller room operation. The real-time coupled operation status includes the media transmission association, power matching relationship and load rate linkage coupling relationship between the entire link equipment. Define the medium transmission coupling coefficient, power matching coupling coefficient, and load rate linkage coupling coefficient, all of which are dimensionless coefficients; and define the parameter abnormality threshold, downstream equipment parameter over-limit threshold, and computer room system energy efficiency benchmark threshold for each device. Each adjustment action in the initial adjustment scheme is decomposed into a single action to obtain the control object, control amplitude, and initial control value of each adjustment action. The control amplitude is the amount of change in the control object parameter caused by the adjustment action. For each disassembled single adjustment action, under the restored real-time coupled operation state, the chain failure path is deduced one by one based on the corresponding coupling relationship and coupling coefficient. The specific deduction logic is as follows: determine whether the node equipment caused by the adjustment action has abnormal parameters, determine whether the abnormal node equipment has caused the downstream equipment parameters to exceed the limit, determine whether the downstream equipment parameters exceed the limit and cause the system energy efficiency to be abnormal, and characterize the quantified chain failure path according to the above levels. Based on the cascading failure paths derived from the deduction, the probability of occurrence of each cascading failure path is quantitatively calculated. Specifically, this quantification is achieved by integrating multi-dimensional correlation parameters and using a weighted summation method. The correlation parameters include the number of node devices in the failure path, the core weight coefficient of each node device, the basic probability of abnormal operating parameters caused by adjustment actions of each node device, and the influence coefficient of the adjustment amplitude corresponding to the adjustment action. Among them, the core weight coefficient is set according to the core priority of each node device in the entire link of the chiller room, and the sum of the core weight coefficients of all node devices is 1. The basic probability of equipment abnormality is determined by the deviation of the real-time operating load rate of the node device from the optimal operating load rate, where the optimal operating load rate is the real-time operating load rate when the device is in its optimal operating state and has the highest energy efficiency. The influence coefficient of the adjustment amplitude is determined by the ratio of the adjustment amplitude of the adjustment action to the rated operating parameter of the corresponding controlled object.

[0009] Furthermore, S3 includes the following: A preset dynamic failure probability threshold is set, which is combined with the redundancy rate of the cooling room equipment, the average failure rate of the entire link equipment, the energy efficiency benchmark threshold of the computer room system, and the energy efficiency normalization coefficient. The calculated probability of occurrence of each cascading failure path is compared with the dynamic failure probability threshold. All adjustment contents corresponding to failure paths whose probability of occurrence exceeds the dynamic failure probability threshold are removed, and the adjustment contents corresponding to failure paths whose probability of occurrence does not exceed the dynamic failure probability threshold are retained to form an initial optimization scheme. The initial optimization scheme is subjected to synergy verification. The synergy adaptation coefficient between each adjustment action is calculated based on the ratio of the adjustment amplitude of each adjustment action to its rated operating parameter and the coupling correlation coefficient between the corresponding control devices. The synergy adaptation is then determined in conjunction with a preset synergy adaptation threshold. If the synergy adaptation coefficient does not reach the preset threshold, the individual synergy contribution of each adjustment action is obtained by averaging the pairwise synergy adaptation values ​​of a single adjustment action with all other adjustment actions. The adjustment action with the lowest individual synergy contribution is deleted, and an alternative adjustment action with the best synergy with the remaining adjustment actions is added. The alternative adjustment action has the same control object as the deleted adjustment action and has the highest average individual synergy contribution with the remaining adjustment actions, with no synergy conflict. The synergy verification is repeated until the synergy adaptation coefficient reaches the preset threshold to obtain the target adjustment scheme. The full-link failure paths and their corresponding probability patterns obtained from the deduction are transformed into dynamic correction parameters for the model. The core parameters of the dynamic correction and control model for the chiller room are updated. The core parameters include the medium transmission coupling coefficient, power matching coupling coefficient, load rate linkage coupling coefficient, and equipment load rate anomaly sensitivity coefficient. The update process combines the probability patterns of all failure paths and the load rate deviation of the core node equipment in the failure paths to achieve adaptive optimization of the model parameters. The updated core parameters are substituted into the dynamic correction and control model of the chiller room. The model correction effect is verified by comparing the predicted system energy efficiency ratio after the model update with the energy efficiency benchmark threshold of the chiller room system. After the verification is passed, the dynamic correction of the model is completed.

[0010] Furthermore, S4 includes the following: Each adjustment action in the target adjustment scheme is standardized and converted into standardized control instructions that can be executed by the equipment. During the conversion process, the control range of the adjustment action is modified in combination with the equipment coupling relationship corresponding to the adjustment action, and the modified control range is limited to not exceeding the preset range of the rated operating parameters of the controlled object. The preset range is set based on the safe operation requirements of the computer room equipment and the energy efficiency optimization target. Based on the coupling relationship between devices, all standardized control commands are time-coordinated and arranged. The execution time interval of commands between any two associated devices is calculated, and the execution commands of each device are issued sequentially according to the time interval to realize the time-sequential control of commands. The standardized and time-sequential execution instructions are output to the corresponding actuators in the refrigeration room, driving the chiller, cooling tower, cooling pump, and chilled pump to complete the corresponding control actions according to the corrected control range and timing. The system collects real-time operating data of all equipment in the data center after the command is executed, calculates the real-time comprehensive energy efficiency ratio of the system and the average real-time operating load rate of the core equipment after execution. The core equipment refers to the key equipment in the entire cooling data center chain, including chillers and cooling towers. The system determines the execution effect by comparing the real-time comprehensive energy efficiency ratio of the system after execution with the data center system energy efficiency benchmark threshold, and the average real-time operating load rate of the core equipment after execution with the optimal operating load rate of the equipment. If the execution effect verification passes, the intelligent management and control of the chiller room is completed; if the execution effect verification fails, the execution deviation data will be fed back to the dynamic correction and control model of the chiller room as input data for the next round of failure path deduction and model dynamic correction, forming a closed loop of management and control.

[0011] A dynamic correction-based intelligent management and control system for a chiller room includes: a data acquisition module, a model management and control module, a scheme optimization module, and an instruction execution module; The data acquisition module collects real-time operating data of all equipment in the data center and the total cooling capacity of the system, verifies the validity of the data, generates a standardized data center operation dataset, and outputs it to the model management module. The model control module has a built-in dynamic correction control model, which is used to receive standardized datasets, restore the real-time coupled operating status of the equipment, analyze the operating status and generate an initial adjustment plan, deduce the chain failure path corresponding to each adjustment action and quantify the failure probability, and output relevant data to the plan optimization module. The scheme optimization module presets a dynamic failure probability threshold, compares the failure probability and removes high-risk adjustment content, performs collaborative verification and iterative optimization on the initial adjustment scheme to obtain the target adjustment scheme, and updates and corrects the core parameters of the control model based on the failure path and probability law, and synchronizes them to the model control module. The instruction execution module transforms the target adjustment plan into standardized control instructions, completes the instruction timing coordination and arrangement, and sends them to the execution mechanism to drive the equipment to run; it collects the running data after execution to determine the execution effect, and feeds back the deviation data to the model control module to form a control closed loop.

[0012] Furthermore, the data acquisition module includes a data acquisition unit and a data verification unit; The data acquisition unit collects real-time operating data of all equipment in the computer room and the total cooling capacity of the system; The data verification unit filters the raw data, removes abnormal and missing data, and generates a standardized data center operation dataset.

[0013] Furthermore, the model control module includes a state restoration unit and a failure simulation unit; The state restoration unit combines the coupling coefficient to restore the real-time coupled operating state of the equipment, analyzes the operating state, and generates an initial adjustment plan; The failure simulation unit is disassembled and adjusted to simulate cascading failure paths and the failure probability is quantified using a weighted summation method.

[0014] Furthermore, the scheme optimization module includes a threshold control unit and a model calibration unit; The threshold control unit presets a dynamic failure probability threshold, eliminates adjustment content with failure probability exceeding the threshold, and optimizes the adjustment actions through collaborative verification and iteration to obtain the target adjustment plan. The model calibration unit updates the core parameters of the model based on the failure path and probability law, verifies the calibration effect, and updates the relevant parameters simultaneously.

[0015] Furthermore, the instruction execution module includes an instruction conversion unit and an execution verification unit; The instruction conversion unit corrects and adjusts the control action range, completes the timing arrangement of control instructions, and issues them for execution. The execution verification unit calculates the system energy efficiency and equipment load rate after execution, determines the execution effect, and feeds back the deviation data to the model control module.

[0016] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention collects and verifies the operational data of all equipment in the refrigeration room through a data acquisition module, providing accurate and reliable data support for intelligent management and control, and avoiding the impact of abnormal data on management accuracy. It uses a model management module to recreate the real-time coupled operating state of the equipment, deduce the chain failure paths corresponding to each adjustment action, and quantify the probability, achieving pre-judgment of adjustment risks and solving the pain point of blind adjustment in existing technologies that easily leads to system anomalies. The scheme optimization module presets a dynamic failure probability threshold, eliminates adjustment content with failure probabilities exceeding the threshold, and iteratively optimizes adjustment actions through collaborative verification to obtain the optimal target adjustment scheme, avoiding adjustment action conflicts. Furthermore, it dynamically updates the core parameters of the model based on failure patterns, completing model correction and improving the predictive accuracy and adaptability of the management model. The instruction execution module transforms the target adjustment scheme into standardized timing instructions, avoiding conflicts in the synchronous control of multiple devices, and forms a closed-loop feedback with execution verification, ensuring that the management effect meets the standards. This invention enables dynamic intelligent management and control of the entire chiller room chain, effectively improving system operating efficiency and stability, reducing equipment failure rate, lowering maintenance costs, adapting to the management and control needs of chiller rooms under different operating conditions, and is highly practical. It solves the technical problems of low accuracy, lack of pre-emptive prevention and control, fixed models, and poor coordination in existing chiller room management and control systems. Attached Figure Description

[0017] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is a schematic diagram of a module of an intelligent control system for a refrigeration room based on dynamic correction, according to the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] Please see Figure 1 The present invention provides the following technical solution: A dynamic correction-based intelligent management and control system for a chiller room includes: a data acquisition module, a model management and control module, a scheme optimization module, and an instruction execution module; The data acquisition module collects real-time operating data of all equipment in the data center and the total cooling capacity of the system, verifies the validity of the data, generates a standardized data center operation dataset, and outputs it to the model management module. The model control module has a built-in dynamic correction control model, which is used to receive standardized datasets, restore the real-time coupled operating status of the equipment, analyze the operating status and generate an initial adjustment plan, deduce the chain failure path corresponding to each adjustment action and quantify the failure probability, and output relevant data to the plan optimization module. The scheme optimization module presets a dynamic failure probability threshold, compares the failure probability and removes high-risk adjustment content, performs collaborative verification and iterative optimization on the initial adjustment scheme to obtain the target adjustment scheme, and updates and corrects the core parameters of the control model based on the failure path and probability law, and synchronizes them to the model control module. The instruction execution module transforms the target adjustment plan into standardized control instructions, completes the instruction timing coordination and arrangement, and sends them to the execution mechanism to drive the equipment to run; it collects the running data after execution to determine the execution effect, and feeds back the deviation data to the model control module to form a control closed loop.

[0020] The data acquisition module includes a data acquisition unit and a data verification unit; The data acquisition unit collects real-time operating data of all equipment in the computer room and the total cooling capacity of the system; The data verification unit filters the raw data, removes abnormal and missing data, and generates a standardized data center operation dataset.

[0021] The model control module includes a state restoration unit and a failure simulation unit; The state restoration unit combines the coupling coefficient to restore the real-time coupled operating state of the equipment, analyzes the operating state, and generates an initial adjustment plan; The failure simulation unit is disassembled and adjusted to simulate cascading failure paths and the failure probability is quantified using a weighted summation method.

[0022] The scheme optimization module includes a threshold control unit and a model calibration unit; The threshold control unit presets a dynamic failure probability threshold, eliminates adjustment content with failure probability exceeding the threshold, and optimizes the adjustment actions through collaborative verification and iteration to obtain the target adjustment plan. The model calibration unit updates the core parameters of the model based on the failure path and probability law, verifies the calibration effect, and updates the relevant parameters simultaneously.

[0023] The instruction execution module includes an instruction conversion unit and an execution verification unit; The instruction conversion unit corrects and adjusts the control action range, completes the timing arrangement of control instructions, and issues them for execution. The execution verification unit calculates the system energy efficiency and equipment load rate after execution, determines the execution effect, and feeds back the deviation data to the model control module.

[0024] A method for intelligent management and control of a chiller room based on dynamic correction includes the following steps: S1. Collect real-time operating data of all equipment in the chiller room, input the real-time operating data into the chiller room dynamic correction and control model for operating status analysis, and generate an initial adjustment plan for the equipment in the chiller room based on the analysis results; S2. In the dynamic correction and control model of the cooling room, the existing operating status of the room is restored by combining real-time operating data. For each adjustment action in the initial adjustment plan, the failure path caused by the chain reaction between the devices in the entire link of the room after its execution is deduced one by one, and the probability of occurrence of each failure path is quantified. S3. Preset a failure probability threshold, remove the adjustment content corresponding to failure paths whose occurrence probability exceeds the failure probability threshold, iteratively optimize the initial adjustment plan to obtain the target adjustment plan; use the failure paths of the entire link and the corresponding probability rules as the basis for dynamic correction, update the parameters of the dynamic correction and control model of the chiller room, and complete the dynamic correction of the model. S4. Transform the target adjustment plan into standardized equipment control commands, output the equipment control commands to the actuators in the refrigeration room, drive the corresponding equipment in the room to perform control actions, and realize intelligent management and control of the refrigeration room.

[0025] S1 includes the following: Real-time operating data of all equipment in the refrigeration room is collected. The equipment in the entire chain includes chillers, cooling towers, cooling pumps, and chilled pumps. The real-time operating data includes the operating power, medium temperature, medium flow rate, and number of devices in operation for each device. At the same time, the total cooling capacity of the computer room system is collected. The validity of the collected real-time operation data is verified by using the parameter fluctuation threshold verification method commonly used in the field of refrigeration room. Abnormal data and missing data that exceed the normal fluctuation range are removed to obtain a standardized data set of refrigeration room operation data, which is then input into the dynamic correction and control model of refrigeration room. The dynamic correction and control model for the chiller room analyzes the operating status based on the data set of the chiller room operation, and calculates the real-time comprehensive energy efficiency ratio of the chiller room system and the real-time operating load rate of each device in sequence. The real-time operating parameters of the chiller, cooling pump and chilled pump are the medium flow rate, and the real-time operating parameters of the cooling tower are the water treatment volume. The dynamic calibration and control model for the chiller room aims to maximize the real-time comprehensive energy efficiency ratio of the chiller room system. It combines the real-time operating load rate of each device and the device linkage operation rules to match the cooling capacity requirements of the chiller room and generate an initial adjustment plan for the joint control of single or multiple devices in the chiller room. The adjustment actions of the initial adjustment plan include adjusting the device frequency, increasing or decreasing the number of devices in operation, and fine-tuning the device operating parameters. The dynamic calibration and control model for the chiller room is a preset simulation analysis model adapted to the operation and control of the entire chain of devices in the chiller room.

[0026] In this embodiment, real-time operating data of all equipment in the refrigeration room is collected. The equipment in the entire chain includes chillers, cooling towers, cooling pumps, and chilled pumps. The real-time operating data includes the operating power, medium temperature, medium flow rate, and number of devices in operation for each device. The total cooling capacity Q of the computer room system is also collected. The validity of the collected real-time operation data is verified by using the parameter fluctuation threshold verification method commonly used in the field of refrigeration room. Abnormal data and missing data that exceed the normal fluctuation range of ±10% are removed to obtain a standardized data center operation dataset. The data center operation dataset is then input into the dynamic correction and control model of the refrigeration room. The dynamic correction and control model for the chiller room analyzes the operating status based on the chiller room operation dataset, sequentially calculating the real-time comprehensive energy efficiency ratio (EER) of the chiller room system and the real-time operating load rate (β) of each device. The real-time comprehensive energy efficiency ratio of the chiller room system is calculated as: EER = Q / ∑Pd, where Pd is the real-time operating power of the d-th device in the entire chain, and ∑Pd is the sum of the real-time operating power of all devices in the entire chain. The real-time operating load rate of each device is calculated as: β = Xs / Xe, where Xs is the real-time operating parameter of the device, and Xe is the rated operating parameter of the device. The real-time operating parameters for the chiller, cooling pump, and refrigeration pump are the medium flow rates, and the real-time operating parameter for the cooling tower is the water treatment capacity. The dynamic correction and control model for the chiller room aims to maximize the real-time comprehensive energy efficiency ratio (EER) of the chiller room system. It combines the real-time operating load rate (β) of each device with the device linkage operation rules, such as device start-stop linkage control and frequency coordinated adjustment, to match the chiller room's cooling capacity requirements. It generates initial adjustment schemes for single or multiple devices in the chiller room, including adjustments to device frequency, increases or decreases in the number of devices in operation, and fine-tuning of device operating parameters. The dynamic correction and control model for the chiller room in this invention is a simulation analysis model adapted to the full-link equipment operation control of the chiller room. It can receive chiller room operating data, analyze operating status, calculate parameters such as energy efficiency or load rate, and generate adjustment schemes. It also supports dynamic correction functions such as chiller room operating status restoration, failure path deduction, and model parameter updates. Its specific model architecture can adopt conventional chiller room simulation models in the field without additional modifications to adapt to the control method of this invention.

[0027] S2 includes the following: In the dynamic correction and control model of the chiller room, the real-time coupled operation status of the entire link equipment in the chiller room is restored by combining the standardized data set of the chiller room operation. The real-time coupled operation status includes the media transmission association, power matching relationship and load rate linkage coupling relationship between the entire link equipment. Define the medium transmission coupling coefficient, power matching coupling coefficient, and load rate linkage coupling coefficient, all of which are dimensionless coefficients; and define the parameter abnormality threshold, downstream equipment parameter over-limit threshold, and computer room system energy efficiency benchmark threshold for each device. Each adjustment action in the initial adjustment scheme is decomposed into a single action to obtain the control object, control amplitude, and initial control value of each adjustment action. The control amplitude is the amount of change in the control object parameter caused by the adjustment action. For each disassembled single adjustment action, under the restored real-time coupled operation state, the chain failure path is deduced one by one based on the corresponding coupling relationship and coupling coefficient. The specific deduction logic is as follows: determine whether the node equipment caused by the adjustment action has abnormal parameters, determine whether the abnormal node equipment has caused the downstream equipment parameters to exceed the limit, determine whether the downstream equipment parameters exceed the limit and cause the system energy efficiency to be abnormal, and characterize the quantified chain failure path according to the above levels. Based on the cascading failure paths derived from the deduction, the probability of occurrence of each cascading failure path is quantitatively calculated. Specifically, this quantification is achieved by integrating multi-dimensional correlation parameters and using a weighted summation method. The correlation parameters include the number of node devices in the failure path, the core weight coefficient of each node device, the basic probability of abnormal operating parameters caused by adjustment actions of each node device, and the influence coefficient of the adjustment amplitude corresponding to the adjustment action. Among them, the core weight coefficient is set according to the core priority of each node device in the entire link of the chiller room, and the sum of the core weight coefficients of all node devices is 1. The basic probability of equipment abnormality is determined by the deviation of the real-time operating load rate of the node device from the optimal operating load rate, where the optimal operating load rate is the real-time operating load rate when the device is in its optimal operating state and has the highest energy efficiency. The influence coefficient of the adjustment amplitude is determined by the ratio of the adjustment amplitude of the adjustment action to the rated operating parameter of the corresponding controlled object.

[0028] In this embodiment, the dynamic correction and control model for the chiller room, combined with the standardized data set of the chiller room operation, restores the real-time coupled operation status of all devices in the chiller room. The real-time coupled operation status includes the media transmission correlation, power matching relationship, and load rate linkage coupling relationship between all devices in the entire link. The media transmission coupling coefficient kq, the power matching coupling coefficient kp, and the load rate linkage coupling coefficient kβ are defined, and kq, kp, and kβ are all dimensionless coefficients. The abnormal parameter threshold Xy of each device, the downstream device parameter over-limit threshold Xa, and the energy efficiency benchmark threshold EERb of the chiller room system are also defined. Each adjustment action in the initial adjustment scheme is decomposed into a single action to obtain the control object, control amplitude Δx and initial control value Xc of each adjustment action. The control amplitude Δx is the change in the control object parameter caused by the adjustment action. For each individual adjustment action after disassembly, under the restored real-time coupled operating state, the cascading failure path is deduced one by one based on the corresponding coupling relationship and coupling coefficient, specifically as follows: Calculate the parameter change ΔXk of the node equipment caused by the adjustment action, and ΔXk=Δx×kt, where kt takes the values ​​kq, kp, and kβ; determine whether the node equipment parameter is abnormal by using |Xc+ΔXk|>Xy; calculate the parameter change ΔXx of the downstream equipment linkage caused by the node abnormality, and ΔXx=ΔXk×kt, and determine whether the downstream equipment parameter exceeds the limit by using |Xcx+ΔXx|>Xa, where Xcx is the initial operating parameter of the downstream equipment; calculate the system energy efficiency ratio EERa after exceeding the limit, and EERa=Q / (∑Pd+ΔP), where ΔP=ΔXx×μp, where μp is the average power conversion coefficient of the entire link, used to convert the parameter change ΔXx of the downstream equipment into the power change ΔP of the entire link in the data center; determine whether the system energy efficiency is abnormal by using |EERa-EERb|>0.1×EERb, and characterize the quantified cascading failure path according to the hierarchy; Based on the cascading failure paths derived from the deduction, the probability of occurrence of each cascading failure path is quantified and calculated. The formula for calculating the probability of occurrence of each failure path is: P = ∑ k∈[1,n] (wk×Pk×K), where P is the overall probability of occurrence of a single failure path, n is the number of node devices in the failure path, and wk is the core weight coefficient of the k-th node device in the failure path, and ∑ k∈[1,n] wk=1, Pk is the base probability that the k-th node device will cause abnormal operating parameters due to adjustment actions, and K is the influence coefficient of the adjustment action's control amplitude; where wk is set according to the coreness of the node device in the entire chiller room chain, for example, wk value is 0.4-0.5 for chillers, wk value is 0.2-0.3 for cooling towers, and wk value is 0.1-0.2 for cooling pumps and chilled pumps; Pk is calculated from the real-time operating load rate of the node device, and the calculation formula is: Pk=1-e -α×|βk-β0| Where βk is the real-time operating load rate of the k-th node device, β0 is the optimal operating load rate, α is the abnormal sensitivity coefficient of the device load rate, and K is calculated by the ratio of the adjustment action amplitude to the rated parameter of the controlled object, i.e., K=1+Δx / Xe, where Δx is the adjustment action amplitude of the controlled object parameter, and Xe is the rated operating parameter of the controlled object.

[0029] S3 includes the following: A preset dynamic failure probability threshold is set, which is combined with the redundancy rate of the cooling room equipment, the average failure rate of the entire link equipment, the energy efficiency benchmark threshold of the computer room system, and the energy efficiency normalization coefficient. The calculated probability of occurrence of each cascading failure path is compared with the dynamic failure probability threshold. All adjustment contents corresponding to failure paths whose probability of occurrence exceeds the dynamic failure probability threshold are removed, and the adjustment contents corresponding to failure paths whose probability of occurrence does not exceed the dynamic failure probability threshold are retained to form an initial optimization scheme. The initial optimization scheme is subjected to synergy verification. The synergy adaptation coefficient between each adjustment action is calculated based on the ratio of the adjustment amplitude of each adjustment action to its rated operating parameter and the coupling correlation coefficient between the corresponding control devices. The synergy adaptation is then determined in conjunction with a preset synergy adaptation threshold. If the synergy adaptation coefficient does not reach the preset threshold, the individual synergy contribution of each adjustment action is obtained by averaging the pairwise synergy adaptation values ​​of a single adjustment action with all other adjustment actions. The adjustment action with the lowest individual synergy contribution is deleted, and an alternative adjustment action with the best synergy with the remaining adjustment actions is added. The alternative adjustment action has the same control object as the deleted adjustment action and has the highest average individual synergy contribution with the remaining adjustment actions, with no synergy conflict. The synergy verification is repeated until the synergy adaptation coefficient reaches the preset threshold to obtain the target adjustment scheme. The full-link failure paths and their corresponding probability patterns obtained from the deduction are transformed into dynamic correction parameters for the model. The core parameters of the dynamic correction and control model for the chiller room are updated. The core parameters include the medium transmission coupling coefficient, power matching coupling coefficient, load rate linkage coupling coefficient, and equipment load rate anomaly sensitivity coefficient. The update process combines the probability patterns of all failure paths and the load rate deviation of the core node equipment in the failure paths to achieve adaptive optimization of the model parameters. The updated core parameters are substituted into the dynamic correction and control model of the chiller room. The model correction effect is verified by comparing the predicted system energy efficiency ratio after the model update with the energy efficiency benchmark threshold of the chiller room system. After the verification is passed, the dynamic correction of the model is completed.

[0030] In this embodiment, a dynamic failure probability threshold P0 is preset, and P0 = γ × λ × (1 - EERb / G), where γ is the redundancy rate of the chiller room equipment, λ is the average failure rate of the entire link equipment, and G is the energy efficiency normalization coefficient. The calculated probability P of each cascading failure path is compared with the dynamic failure probability threshold P0. All adjustment contents corresponding to failure paths with P > P0 are removed, and the adjustment contents corresponding to failure paths with P ≤ P0 are retained to form an initial optimization scheme. The initial optimization scheme is subjected to a coordination verification, and the coordination adaptation coefficient η between the retained adaptation actions is calculated, where η = (1 / m)∑ i，j∈[1,m],i≠j(1-|(Δxi / Xei)-(Δxj / Xej)|×kij), where m is the number of adjustment actions retained in the initial optimization scheme, Δxi and Δxj are the control amplitudes of the i-th and j-th adjustment actions, respectively, Xei and Xej are the rated operating parameters of the control objects of the i-th and j-th adjustment actions, respectively, and kij is the coupling correlation coefficient between the control objects of the i-th and j-th adjustment actions, kq, kp, and kβ are selected according to the equipment coupling relationship; when the coordination adaptation coefficient η is greater than or equal to the preset threshold η0, it is judged as coordination adaptation, and the corresponding initial optimization scheme is marked as the target adjustment scheme; if the coordination adaptation coefficient η is less than the preset threshold η0, the single coordination contribution Ci of each adjustment action is calculated, and Ci=[1 / (m-1)]×∑ j≠i (1-|Δxi / Xei)-(Δxj / Xej)|×kij), where, ∑ j≠i To sum all actions j except i, the adaptation action with the smallest single collaborative contribution Ci is deleted, and the alternative adaptation action with the largest average single collaborative contribution among the remaining adaptation actions is added. η is recalculated until the collaborative adaptation coefficient η is greater than or equal to the preset threshold η0, and the target adaptation scheme is obtained. The derived end-to-end failure paths and their corresponding probability patterns are transformed into dynamic correction parameters for the model. This updates the core parameters of the dynamic correction and control model for the chiller room, specifically: The update formula for coupling coefficients kq, kp, and kβ is k'=k×[1+(1 / h)∑ f∈[1,h] The update formula for the abnormal sensitivity coefficient α of equipment load rate is α'=α×(1+(1 / h)∑ f∈[1,h] |βkf-β0|×Pf), where k' is the updated coupling coefficient, k is the original coupling coefficient, h is the total number of failure paths in the entire link, Pf is the comprehensive occurrence probability of the f-th failure path, and βkf is the real-time operating load rate of the core node equipment in the f-th failure path; the update of the equipment load rate anomaly sensitivity coefficient α is based on the failure paths in the entire link: the load rate deviation of the core node equipment in the failure path |βkf-β0| characterizes the correlation between failure and load rate anomaly, and the failure path occurrence probability Pf is used as the weight to characterize the credibility of the failure law. The averaging process ensures that the correction amplitude is stable, so that α increases adaptively with the actual failure law, improves the model's sensitivity to the anomaly caused by load rate deviation, and realizes the closed-loop improvement of dynamic correction and prediction accuracy.

[0031] The updated core parameters are substituted into the dynamic correction and control model of the chiller room, and the model correction is verified by |EERy-EERb|≤0.05×EERb, where EERy is the system energy efficiency ratio predicted based on the current operating data after the model update; after the verification is passed, the dynamic correction of the model is completed.

[0032] Furthermore, S4 includes the following: Each adjustment action in the target adjustment scheme is standardized and converted into standardized control instructions that can be executed by the equipment. During the conversion process, the control range of the adjustment action is modified in combination with the equipment coupling relationship corresponding to the adjustment action, and the modified control range is limited to not exceeding the preset range of the rated operating parameters of the controlled object. The preset range is set based on the safe operation requirements of the computer room equipment and the energy efficiency optimization target. Based on the coupling relationship between devices, all standardized control commands are time-coordinated and arranged. The execution time interval of commands between any two associated devices is calculated, and the execution commands of each device are issued sequentially according to the time interval to realize the time-sequential control of commands. The standardized and time-sequential execution instructions are output to the corresponding actuators in the refrigeration room, driving the chiller, cooling tower, cooling pump, and chilled pump to complete the corresponding control actions according to the corrected control range and timing. The system collects real-time operating data of all equipment in the data center after the command is executed, calculates the real-time comprehensive energy efficiency ratio of the system and the average real-time operating load rate of the core equipment after execution. The core equipment refers to the key equipment in the entire cooling data center chain, including chillers and cooling towers. The system determines the execution effect by comparing the real-time comprehensive energy efficiency ratio of the system after execution with the data center system energy efficiency benchmark threshold, and the average real-time operating load rate of the core equipment after execution with the optimal operating load rate of the equipment. If the execution effect verification passes, the intelligent management and control of the chiller room is completed; if the execution effect verification fails, the execution deviation data will be fed back to the dynamic correction and control model of the chiller room as input data for the next round of failure path deduction and model dynamic correction, forming a closed loop of management and control.

[0033] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0034] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for intelligent management and control of a chiller room based on dynamic correction, characterized in that: The method includes the following steps: S1. Collect real-time operating data of all equipment in the chiller room, input the real-time operating data into the chiller room dynamic correction and control model for operating status analysis, and generate an initial adjustment plan for the equipment in the chiller room based on the analysis results; S2. In the dynamic correction and control model of the cooling room, the existing operating status of the room is restored by combining real-time operating data. For each adjustment action in the initial adjustment plan, the failure path caused by the chain reaction between the devices in the entire link of the room after its execution is deduced one by one, and the probability of occurrence of each failure path is quantified. S3. Preset a failure probability threshold, remove the adjustment content corresponding to failure paths whose occurrence probability exceeds the failure probability threshold, iteratively optimize the initial adjustment plan to obtain the target adjustment plan; use the failure paths of the entire link and the corresponding probability rules as the basis for dynamic correction, update the parameters of the dynamic correction and control model of the chiller room, and complete the dynamic correction of the model. S4. Transform the target adjustment plan into standardized equipment control commands, output the equipment control commands to the actuators in the refrigeration room, drive the corresponding equipment in the room to perform control actions, and realize intelligent management and control of the refrigeration room.

2. The intelligent control method for a chiller room based on dynamic correction according to claim 1, characterized in that: S1 includes the following: Real-time operating data of all equipment in the refrigeration room is collected. The equipment in the entire chain includes chillers, cooling towers, cooling pumps, and chilled pumps. The real-time operating data includes the operating power, medium temperature, medium flow rate, and number of devices in operation for each device. At the same time, the total cooling capacity of the computer room system is collected. The validity of the collected real-time operation data is verified by using the parameter fluctuation threshold verification method commonly used in the field of refrigeration room. Abnormal data and missing data that exceed the normal fluctuation range are removed to obtain a standardized data set of refrigeration room operation data, which is then input into the dynamic correction and control model of refrigeration room. The dynamic correction and control model for the chiller room analyzes the operating status based on the data set of the chiller room operation, and calculates the real-time comprehensive energy efficiency ratio of the chiller room system and the real-time operating load rate of each device in sequence. The real-time operating parameters of the chiller, cooling pump and chilled pump are the medium flow rate, and the real-time operating parameters of the cooling tower are the water treatment volume. The dynamic calibration and control model for the chiller room aims to maximize the real-time comprehensive energy efficiency ratio of the chiller room system. It combines the real-time operating load rate of each device and the device linkage operation rules to match the cooling capacity requirements of the chiller room and generate an initial adjustment plan for the joint control of single or multiple devices in the chiller room. The adjustment actions of the initial adjustment plan include adjusting the device frequency, increasing or decreasing the number of devices in operation, and fine-tuning the device operating parameters. The dynamic calibration and control model for the chiller room is a preset simulation analysis model adapted to the operation and control of the entire chain of devices in the chiller room.

3. The intelligent management and control method for a chiller room based on dynamic correction according to claim 2, characterized in that: S2 includes the following: In the dynamic correction and control model of the chiller room, the real-time coupled operation status of the entire link equipment in the chiller room is restored by combining the standardized data set of the chiller room operation. The real-time coupled operation status includes the media transmission association, power matching relationship and load rate linkage coupling relationship between the entire link equipment. Define the medium transmission coupling coefficient, power matching coupling coefficient, and load rate linkage coupling coefficient, all of which are dimensionless coefficients. And define the abnormal parameter thresholds for each device, the over-limit thresholds for downstream device parameters, and the energy efficiency benchmark thresholds for the computer room system; Each adjustment action in the initial adjustment scheme is decomposed into a single action to obtain the control object, control amplitude, and initial control value of each adjustment action. The control amplitude is the amount of change in the control object parameter caused by the adjustment action. For each disassembled single adjustment action, under the restored real-time coupled operation state, the chain failure path is deduced one by one based on the corresponding coupling relationship and coupling coefficient. The specific deduction logic is as follows: determine whether the node equipment caused by the adjustment action has abnormal parameters, determine whether the abnormal node equipment has caused the downstream equipment parameters to exceed the limit, determine whether the downstream equipment parameters exceed the limit and cause the system energy efficiency to be abnormal, and characterize the quantified chain failure path according to the above levels. Based on the cascading failure paths derived from the deduction, the probability of occurrence of each cascading failure path is quantitatively calculated. Specifically, this quantification is achieved by integrating multi-dimensional correlation parameters and using a weighted summation method. The correlation parameters include the number of node devices in the failure path, the core weight coefficient of each node device, the basic probability of abnormal operating parameters caused by adjustment actions of each node device, and the influence coefficient of the adjustment amplitude corresponding to the adjustment action. Among them, the core weight coefficient is set according to the core priority of each node device in the entire link of the chiller room, and the sum of the core weight coefficients of all node devices is 1. The basic probability of equipment abnormality is determined by the deviation of the real-time operating load rate of the node device from the optimal operating load rate, where the optimal operating load rate is the real-time operating load rate when the device is in its optimal operating state and has the highest energy efficiency. The influence coefficient of the adjustment amplitude is determined by the ratio of the adjustment amplitude of the adjustment action to the rated operating parameter of the corresponding controlled object.

4. The intelligent control method for a chiller room based on dynamic correction according to claim 3, characterized in that: S3 includes the following: A preset dynamic failure probability threshold is set, which is combined with the redundancy rate of the cooling room equipment, the average failure rate of the entire link equipment, the energy efficiency benchmark threshold of the computer room system, and the energy efficiency normalization coefficient. The calculated probability of occurrence of each cascading failure path is compared with the dynamic failure probability threshold. All adjustment contents corresponding to failure paths whose probability of occurrence exceeds the dynamic failure probability threshold are removed, and the adjustment contents corresponding to failure paths whose probability of occurrence does not exceed the dynamic failure probability threshold are retained to form an initial optimization scheme. The initial optimization scheme is checked for synergy. The synergy adaptation coefficient between each adjustment action is calculated based on the ratio of the adjustment amplitude of each adjustment action to its rated operating parameter and the coupling correlation coefficient between the corresponding control devices. The synergy adaptation is determined in combination with the preset synergy adaptation threshold. If the coordination adaptation coefficient does not reach the preset threshold, the individual coordination contribution of each adjustment action is obtained by averaging the pairwise coordination adaptation values ​​of a single adjustment action with all other adjustment actions; the adjustment action with the lowest individual coordination contribution is deleted, and an alternative adjustment action with the best coordination with the remaining adjustment actions is added. The alternative adjustment action has the same control object as the deleted adjustment action, and has the highest average individual coordination contribution with the remaining adjustment actions and no coordination conflict. Repeat the collaboration verification until the collaboration adaptation coefficient reaches the preset threshold to obtain the target adaptation solution. The full-link failure paths and their corresponding probability patterns obtained from the deduction are transformed into dynamic correction parameters for the model. The core parameters of the dynamic correction and control model for the chiller room are updated. The core parameters include the medium transmission coupling coefficient, power matching coupling coefficient, load rate linkage coupling coefficient, and equipment load rate anomaly sensitivity coefficient. The update process combines the probability patterns of all failure paths and the load rate deviation of the core node equipment in the failure paths to achieve adaptive optimization of the model parameters. The updated core parameters are substituted into the dynamic correction and control model of the chiller room. The model correction effect is verified by comparing the predicted system energy efficiency ratio after the model update with the energy efficiency benchmark threshold of the chiller room system. After the verification is passed, the dynamic correction of the model is completed.

5. The intelligent control method for a chiller room based on dynamic correction according to claim 4, characterized in that: S4 includes the following: Each adjustment action in the target adjustment scheme is standardized and converted into standardized control instructions that can be executed by the equipment. During the conversion process, the control range of the adjustment action is modified in combination with the equipment coupling relationship corresponding to the adjustment action, and the modified control range is limited to not exceeding the preset range of the rated operating parameters of the controlled object. The preset range is set based on the safe operation requirements of the computer room equipment and the energy efficiency optimization target. Based on the coupling relationship between devices, all standardized control commands are time-coordinated and arranged. The execution time interval of commands between any two associated devices is calculated, and the execution commands of each device are issued sequentially according to the time interval to realize the time-sequential control of commands. The standardized and time-sequential execution instructions are output to the corresponding actuators in the refrigeration room, driving the chiller, cooling tower, cooling pump, and chilled pump to complete the corresponding control actions according to the corrected control range and timing. The system collects real-time operating data of all equipment in the data center after the command is executed, and calculates the real-time comprehensive energy efficiency ratio of the system and the average real-time operating load rate of the core equipment. The core equipment refers to the key equipment in the entire cooling data center chain, including chillers and cooling towers. The effectiveness of command execution is determined by comparing the real-time comprehensive energy efficiency ratio of the system after execution with the energy efficiency benchmark threshold of the data center system, and the average real-time operating load rate of the core equipment after execution with the optimal operating load rate of the equipment. If the execution effect verification passes, the intelligent management and control of the chiller room is completed; if the execution effect verification fails, the execution deviation data will be fed back to the dynamic correction and control model of the chiller room as input data for the next round of failure path deduction and model dynamic correction, forming a closed loop of management and control.

6. A dynamic correction-based intelligent management and control system for a chiller room, applied to the dynamic correction-based intelligent management and control method for a chiller room according to any one of claims 1-5, characterized in that: The system includes: a data acquisition module, a model management and control module, a scheme optimization module, and an instruction execution module; The data acquisition module collects real-time operating data of all equipment in the data center and the total cooling capacity of the system, verifies the validity of the data, generates a standardized data center operation dataset, and outputs it to the model management module. The model control module has a built-in dynamic correction control model, which is used to receive standardized datasets, restore the real-time coupled operating status of the equipment, analyze the operating status and generate an initial adjustment plan, deduce the chain failure path corresponding to each adjustment action and quantify the failure probability, and output relevant data to the plan optimization module. The scheme optimization module presets a dynamic failure probability threshold, compares the failure probability and removes high-risk adjustment content, performs collaborative verification and iterative optimization on the initial adjustment scheme to obtain the target adjustment scheme, and updates and corrects the core parameters of the control model based on the failure path and probability law, and synchronizes them to the model control module. The instruction execution module transforms the target adjustment scheme into standardized control instructions, completes the instruction timing coordination and arrangement, and sends them to the execution mechanism to drive the equipment to run; it collects the running data after execution to determine the execution effect, and feeds back the deviation data to the model control module to form a control closed loop.

7. The intelligent control system for a chiller room based on dynamic correction according to claim 6, characterized in that: The data acquisition module includes a data acquisition unit and a data verification unit; The data acquisition unit collects real-time operating data of all equipment in the computer room and the total cooling capacity of the system. The data verification unit filters the original data, removes abnormal and missing data, and generates a standardized data center operation dataset.

8. The intelligent control system for a chiller room based on dynamic correction according to claim 6, characterized in that: The model control module includes a state restoration unit and a failure simulation unit; The state restoration unit combines the coupling coefficient to restore the real-time coupled operating state of the equipment, analyzes the operating state, and generates an initial adjustment plan. The failure prediction unit breaks down the adjustment actions, predicts the cascading failure path, and quantifies the failure probability using a weighted summation method.

9. The intelligent control system for a chiller room based on dynamic correction according to claim 6, characterized in that: The scheme optimization module includes a threshold control unit and a model calibration unit; The threshold control unit presets a dynamic failure probability threshold, eliminates adjustment content with failure probability exceeding the threshold, and optimizes the adjustment action through collaborative verification and iteration to obtain the target adjustment scheme. The model correction unit updates the core parameters of the model based on the failure path and probability law, verifies the correction effect, and updates relevant parameters synchronously.

10. The intelligent control system for a chiller room based on dynamic correction according to claim 6, characterized in that: The instruction execution module includes an instruction conversion unit and an execution verification unit; The instruction conversion unit corrects and adjusts the control range of the action, completes the timing arrangement of the control instructions, and issues them for execution. The execution verification unit calculates the system energy efficiency and equipment load rate after execution, determines the execution effect, and feeds back the deviation data to the model control module.

Images