A transformer digital cooling control system and method fusing hotspot prediction

Abstract

The present application relates to the technical field of transformer cooling, and particularly relates to a transformer digital cooling control system and control method fusing hot spot prediction, comprising: a sensing module, which is used for collecting multi-source state data of the transformer in real time; a prediction module, which is used for obtaining a hot spot temperature prediction value of the transformer based on the multi-source state data collected at all sampling times; a control module, which is used for judging whether there is an over-temperature risk based on the hot spot temperature prediction value; when it is judged that there is an over-temperature risk, the obtained hot spot temperature prediction value is used as a feedforward input quantity of a PID controller to generate a cooling control instruction; and an execution module, which is used for driving a cooling device to operate according to the cooling control instruction generated by the control module, wherein when the cooling device is driven, the cooling device to be put into operation is determined according to the priority of the cumulative running time of the cooling device from short to long; the communication time delay is reduced, and the real-time performance and reliability of the cooling control are improved.

Description

Technical Field

[0001] This invention relates to the field of transformer cooling technology, and in particular to a digital cooling control system and method for transformers that integrates hotspot prediction. Background Technology

[0002] As a core piece of equipment in the power system, the operational reliability of large power transformers directly affects the safe and stable operation of the power grid and the continuity of power supply. The cooling system, by adjusting the combination of cooler oil pumps and fans to control hot spot temperatures, is a crucial link in ensuring the long-term safe operation of transformers. With the construction of new power systems and the continuous improvement of the digitalization and intelligence of substations, traditional cooling control methods based on fixed temperature thresholds are no longer sufficient to meet the development needs of frequent load fluctuations, complex operating conditions, and autonomous control. Therefore, how to construct an active cooling control method oriented towards hot spot temperature prediction has become an important direction in digital cooling control research.

[0003] In recent years, scholars both domestically and internationally have conducted extensive research on digital cooling control of transformers, mainly focusing on three aspects: first, hot spot temperature prediction based on thermal models or data-driven models; second, cooling control strategies based on PID, fuzzy control, and intelligent optimization algorithms; and third, condition monitoring and fault diagnosis technologies for cooling equipment. While these studies have made positive progress in improving the intelligence level of cooling control, two shortcomings remain: firstly, most methods still rely primarily on feedback control, lacking coordinated design between hot spot temperature prediction, control decision-making, equipment scheduling, and fault management, making it difficult to fully utilize predictive information; secondly, existing research largely depends on imported controllers or host computer platforms, with relatively insufficient research on engineering applications using domestically produced MCU platforms. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides a digital cooling control system and method for transformers that integrates hotspot prediction. This system collects multi-source data such as top-layer oil temperature, ambient temperature, load current, and the operating status of fans and oil pumps in real time. It also performs preprocessing, hotspot temperature prediction, cooling demand assessment, control decision-making, and fault diagnosis on the collected data. This enables online calculation of cooling control strategies and dynamic adjustment of the combined operation of fans and oil pumps according to control commands, achieving on-demand deployment and intelligent scheduling of cooling equipment.

[0005] In a first aspect, embodiments of the present invention provide a digital cooling control system for transformers that integrates hotspot prediction, comprising: The sensing module is used to collect multi-source status data of the transformer in real time. The multi-source status data includes at least the top oil temperature, ambient temperature and load current. The prediction module obtains the predicted hot spot temperature of the transformer based on the multi-source state data collected at all sampling times; The control module determines whether there is an over-temperature risk based on the hot spot temperature prediction value; when it determines that there is an over-temperature risk, it uses the obtained hot spot temperature prediction value as the feedforward input of the PID controller to generate a cooling control command. The execution module drives the cooling equipment to operate according to the cooling control commands generated by the control module. When driving the cooling equipment, the cooling equipment to be put into operation is determined according to the priority of the cumulative running time of the cooling equipment from shortest to longest.

[0006] Preferably, the method by which the prediction module obtains the predicted hot spot temperature value of the transformer includes: The load factor is obtained from the ratio of load current to rated current in the multi-source status data. The top oil temperature rise is obtained based on the load rate and the rated top oil temperature rise. The winding hot spot temperature rise is obtained based on the load rate and the rated hot spot temperature rise. Based on the obtained top oil temperature rise, winding hot spot temperature rise, and ambient temperature, the predicted value of the hot spot temperature is obtained.

[0007] Preferably, in the control module, the method for generating cooling control commands using the obtained hot spot temperature prediction value as the feedforward input of the PID controller includes: The difference between the predicted hot spot temperature and the preset target hot spot temperature is used as the hot spot temperature control error, and the hot spot temperature control error and its rate of change are used as the input of the PID controller. The PID controller corrects its proportional, integral, and derivative coefficients online according to preset fuzzy inference rules. The PID controller outputs a control quantity based on the corrected proportional coefficient, integral coefficient, and derivative coefficient, and the control quantity constitutes the cooling control command.

[0008] Preferably, the control output of the PID controller includes: ; In the formula: To control the quantity; This is due to the temperature control error of the hot spot. This is the proportionality coefficient; The integral coefficient; These are the differential coefficients; The hot spot temperature control error It is obtained by pre-setting target values ​​and predicted values ​​of hotspot temperatures.

[0009] Preferably, the control quantity Determined in the following manner; Using control variables and hot spot temperature control error Construct an optimization function F; the optimization function F is: ; In the formula: To control performance indicators; and These are the weighting coefficients for the hotspot temperature control error and the control quantity, respectively. Determine the control quantity according to the constructed optimization function F. .

[0010] Preferably, the execution module further includes the following before driving the cooling device: A health evaluation function for each cooling device is constructed based on its device temperature, operating current, and communication status. The constructed health evaluation function is used to evaluate the cooling equipment and select all target cooling equipment. Prioritize all target cooling devices by using the cumulative operating time of each target cooling device.

[0011] Secondly, a digital cooling control method for transformers that integrates hotspot prediction includes: Real-time acquisition of multi-source status data of the transformer, including at least top oil temperature, ambient temperature and load current; The predicted hot spot temperature of the transformer is obtained based on the multi-source state data collected at all sampling times. The hot spot temperature prediction value is used to determine whether there is an over-temperature risk; when it is determined that there is an over-temperature risk, the obtained hot spot temperature prediction value is used as the feedforward input of the PID controller to generate a cooling control command. The cooling equipment is driven to operate according to the cooling control commands generated by the control module. When driving the cooling equipment, the cooling equipment to be put into operation is determined according to the priority of the cumulative running time of the cooling equipment from shortest to longest.

[0012] Preferably, when a cooling device malfunctions, all backup cooling devices are obtained, and the cumulative operating time of all backup cooling devices is used to determine whether the malfunctioning cooling device should be rotated.

[0013] Preferably, the method for determining whether there is an overheating risk based on the hotspot temperature prediction value includes: obtaining the degree of deviation by subtracting the hotspot temperature prediction value from the target hotspot temperature value, and determining whether an overheating risk has occurred based on the obtained degree of deviation.

[0014] Preferably, the method for generating cooling control commands includes: The difference between the predicted hot spot temperature and the target hot spot temperature is used as the control error, and the control error and its rate of change are used as the input of the PID controller. The PID controller corrects its proportional, integral, and derivative coefficients online according to preset fuzzy inference rules. The PID controller outputs a control quantity based on the corrected proportional coefficient, integral coefficient, and derivative coefficient, and the control quantity constitutes the cooling control command.

[0015] The beneficial effects of this invention are as follows: This application realizes an integrated design of operation status acquisition, predictive analysis, control optimization and equipment management; the system adopts a combination of edge computing and station-side collaborative control, and deploys hot spot temperature prediction and control decisions to be completed on the edge side, which improves the real-time performance and reliability of cooling control while reducing communication latency.

[0016] This application constructs a multi-source state perception and hotspot temperature prediction model, introduces cooling demand assessment, predictive-driven control, intelligent scheduling of cooling equipment, and digital fault diagnosis and fault tolerance strategies, forming a closed-loop control system of "state perception - predictive analysis - control decision - intelligent scheduling - reliability assurance", realizing the transformation of cooling control from passive feedback to predictive-driven; effectively improving prediction accuracy, control performance, operating economy and system reliability. Attached Figure Description

[0017] Figure 1 A block diagram of a digital cooling control system for transformers that integrates hotspot prediction, provided as an embodiment of the present invention; Figure 2 This is a schematic diagram of the structural connection of the carbon dioxide refrigerant circulation system in an embodiment of the present invention; Figure 3 This is a diagram illustrating the overall framework of predictive-driven digital cooling control in an embodiment of the present invention. Figure 4 This is a diagram of the hotspot temperature prediction model in an embodiment of the present invention; Figure 5 This is a flowchart of the predictive drive control algorithm in an embodiment of the present invention; Figure 6 This is a flowchart illustrating the intelligent scheduling and fault tolerance process in an embodiment of the present invention. Figure 7 This is a block diagram of the integrated control unit for the cooler in an embodiment of the present invention; Figure 8 This is a schematic diagram of the internal layout of the transformer digital control cabinet in an embodiment of the present invention; Figure 9 A schematic diagram of a real-world scenario in an embodiment of the present invention; Figure 10 This is a comparison curve of the predicted hotspot temperature and the measured value in an embodiment of the present invention; Figure 11 This is a graph showing the temperature change curves of hotspots under different control strategies in an embodiment of the present invention. Figure 12 This is a bar chart showing the distribution of device operating time in an embodiment of the present invention; Figure 13 This is a hotspot temperature curve of the device in an embodiment of the present invention. Detailed Implementation

[0018] To enable those skilled in the art to better understand the technical solutions of the present invention, exemplary embodiments of the present invention are described below in conjunction with the accompanying drawings, including various details of the embodiments of the present invention to aid understanding. These should be considered merely exemplary. Therefore, those skilled in the art should recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present invention. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0019] Where there is no conflict, the various embodiments of the present invention and the features thereof may be combined with each other.

[0020] As used herein, the term “and / or” includes any and all combinations of one or more related enumerated entries.

[0021] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used herein, the singular forms “a” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will also be understood that when the terms “comprising” and / or “made of” are used in this specification, the presence of the stated feature, integral, step, operation, element, and / or component is specified, but the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or groups thereof is not excluded. Terms such as “connected” or “linked” are not limited to physical or mechanical connections but can include electrical connections, whether direct or indirect.

[0022] Unless otherwise specified, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art. It will also be understood that terms such as those defined in commonly used dictionaries should be interpreted as having the meaning consistent with their meaning in the context of the relevant art and the invention, and will not be interpreted as having an idealized or overly formal meaning unless expressly so defined herein.

[0023] In the technical solution of this invention, the collection, storage, use, processing, transmission, provision, and disclosure of user personal information all comply with relevant laws and regulations and do not violate public order and good morals. The use of user data in this technical solution follows relevant national laws and regulations (e.g., the "Information Security Technology - Personal Information Security Specification"). For example: appropriate measures are taken for personal information access control; restrictions are imposed on the display of personal information; the purpose of using personal information does not exceed the scope of direct or reasonable association; and explicit identity targeting is eliminated when using personal information to avoid precisely locating a specific individual. Example 1

[0024] This embodiment provides a digital cooling control system for transformers that integrates hotspot prediction, including: The sensing module is used to collect multi-source status data of the transformer in real time. The multi-source status data includes at least the top oil temperature, ambient temperature and load current. The prediction module obtains the predicted hot spot temperature of the transformer based on the multi-source state data collected at all sampling times; The control module determines whether there is an over-temperature risk based on the hot spot temperature prediction value; when it determines that there is an over-temperature risk, it uses the obtained hot spot temperature prediction value as the feedforward input of the PID controller to generate a cooling control command. The execution module drives the cooling equipment to operate according to the cooling control commands generated by the control module. When driving the cooling equipment, the cooling equipment to be put into operation is determined according to the priority of the cumulative running time of the cooling equipment from shortest to longest.

[0025] In one specific embodiment, this embodiment constructs a predictive-driven digital cooling control system based on a domestically produced MCU. For example... Figure 1 The system adopts an overall architecture of "multi-source state perception - edge intelligent computing - control decision execution - station-end collaborative management". The system consists of a perception layer, an execution layer, an edge computing layer, and a station-end management layer. In the perception layer, multi-source data information such as top oil temperature, ambient temperature, load current, and fan and oil pump operating status are collected in real time through temperature sensors, current transformers, and cooling equipment status detection units. In the edge computing layer, a domestic industrial-grade MCU is used as the core to preprocess the collected data, predict hot spot temperatures, assess cooling demand, make control decisions, and diagnose faults, realizing online calculation of cooling control strategies. In the execution layer, the combined operation of fans and oil pumps is dynamically adjusted according to control commands to realize on-demand deployment and intelligent scheduling of cooling equipment. In the station-end management layer, data interaction is achieved with the integrated automation system through standard industrial communication protocols, providing support for equipment status monitoring, operation management, and remote maintenance.

[0026] The architecture constructed in this embodiment realizes an integrated design of operation status acquisition, predictive analysis, control optimization, and equipment management. The system adopts a combination of edge computing and station-side collaborative control, deploying hotspot temperature prediction and control decisions at the edge, which improves the real-time performance and reliability of cooling control while reducing communication latency, and provides a unified data foundation, real-time computing environment, and control interface for predictive-driven digital cooling control methods.

[0027] The platform in this embodiment uses a domestically produced industrial-grade MCU as its core processor. Compared with traditional PLCs or imported controllers, this processor has advantages such as independent controllability, rich interfaces, flexible development, and low deployment costs, making it more suitable for intelligent control scenarios at the substation edge.

[0028] In this embodiment, to support the real-time operation of the predictive-driven control algorithm, the platform is designed with four functional modules: data acquisition, control calculation, communication management, and device driving. Figure 2 As shown in the diagram, the data acquisition module completes multi-source state fusion, the control decision module executes hotspot temperature prediction and control decisions, the communication management module is responsible for data interaction between field equipment and the station control layer, and the equipment drive module implements control command execution and status feedback, thus forming a complete edge closed-loop control process.

[0029] A digital control platform was established based on domestically produced MCUs, achieving deep integration of multi-source status acquisition, edge intelligent computing, and control execution. The platform has excellent real-time computing capabilities and module expansion capabilities, and can be further integrated with functions such as online monitoring, equipment health assessment, and intelligent operation and maintenance.

[0030] like Figure 3 The embodiment shown provides a predictive-driven digital cooling control framework. Traditional large power transformer cooling systems typically activate fans or oil pumps once the top oil temperature reaches a set threshold. This is a typical feedback control method, which suffers from problems such as response lag, frequent equipment start-ups and shutdowns, and insufficient operational balance. To address these shortcomings, this embodiment utilizes a predictive-driven digital cooling control method based on a domestically produced MCU, constructing a closed-loop control framework encompassing "state perception—hotspot prediction—demand assessment—predictive control—intelligent scheduling—fault tolerance."

[0031] In this embodiment, the system first collects information such as top oil temperature, ambient temperature, load current, and cooling equipment operating status, and performs filtering and outlier processing. Then, it predicts the winding hot spot temperature based on the IEC 60076 and IEEE C57.91 thermal models. Based on the prediction results, it assesses the cooling demand and uses a predictive-driven fuzzy PID algorithm to dynamically adjust the operating status of the fan and oil pump. Finally, it combines the cumulative operating time and health status of the equipment to achieve balanced scheduling of cooling equipment and automatic fault switching, thereby improving the control accuracy, operating economy, and reliability of the cooling system.

[0032] In one embodiment, after collecting multi-source state data of the transformer, a multi-source state data vector is constructed: ; In the formula: A multi-source state data vector; For the top oil temperature rise; The ambient temperature; Load rate; This indicates the operating status of the cooling equipment. In this embodiment The load factor is calculated as follows: ; In the formula: This is the actual load current. This is the rated current.

[0033] After obtaining the load rate Then, the top oil temperature rise was obtained according to the IEC recommended thermal model; the methods for obtaining the top oil temperature rise include: ; In the formula: R is the rated top oil temperature rise; R is the ratio of load loss to no-load loss. This refers to the oil temperature index.

[0034] The winding hot spot temperature rise is obtained based on the load rate and the rated hot spot temperature rise. The specific methods used in the calculation include: ; In the formula: This is the rated hot spot temperature rise; This represents the winding index.

[0035] In this embodiment, after obtaining the top layer oil temperature rise and the winding hot spot temperature rise, the predicted hot spot temperature value is obtained in the following manner: ; In the formula: These are predicted values ​​for hotspot temperatures. Ambient temperature; For the top oil temperature rise; This is due to the temperature rise of the winding hot spot.

[0036] In one embodiment; after obtaining the predicted hotspot temperature value Subsequently, the predicted hotspot temperature values ​​were obtained. As input for cooling demand assessment, it provides a basis for predictive drive control decisions; its hotspot temperature prediction model is shown below. Figure 4 As shown.

[0037] When using the obtained hotspot temperature prediction values ​​as a basis for decision-making, Compare with the target temperature; determine the degree of deviation; specifically: ; in, This is the target value for the hotspot temperature.

[0038] when( When (), it indicates that there is a risk of the hot spot temperature exceeding the limit in the future, and the cooling equipment needs to be started in advance; when ( When this happens, maintain the current operating mode.

[0039] When it is determined that the cooling equipment needs to be started in advance, the obtained hot spot temperature prediction value is used as the feedforward input of the PID controller to generate cooling control commands.

[0040] In one embodiment, the difference between the predicted hotspot temperature and the preset target hotspot temperature is used as the control error, and the control error and its rate of change are used as the input to the PID controller. The PID controller corrects its proportional, integral, and derivative coefficients online according to preset fuzzy inference rules. The PID controller outputs a control quantity based on the corrected proportional coefficient, integral coefficient, and derivative coefficient, and the control quantity constitutes the cooling control command.

[0041] The control outputs of the PID controller include: ; In the formula: To control the quantity; This is due to the temperature control error of the hot spot. This is the proportionality coefficient; The integral coefficient; These are the differential coefficients; The hot spot temperature control error It is obtained through preset hotspot temperature target values ​​and hotspot temperature prediction values, specifically including: ; In the formula: The preset target temperature value for hotspots; Here is the predicted hotspot temperature at time t.

[0042] In one embodiment, the parameters of a traditional PID controller are fixed, and its control performance tends to degrade when the operating conditions change. Therefore, fuzzy control is introduced to achieve online revision of PID parameters. The method for the PID controller to correct its proportional coefficient, integral coefficient, and derivative coefficient online according to preset fuzzy inference rules specifically includes: To control error and error change rate As input to the fuzzy controller, the PID parameter correction amount , and For output, the relationship can be represented as: ; ; ; In the formula: , and These represent the fuzzy inference rules for proportional, integral, and derivative parameters, respectively. The controller adjusts the PID parameters in real time based on the hot spot temperature change trend, enabling the control gain to be dynamically adjusted according to the operating state, thereby improving the system's dynamic response capability and robustness.

[0043] Unlike traditional feedback control methods, this embodiment uses the output of the hot spot temperature prediction model as the feedforward input to the controller. This allows the control system to detect changes in the transformer's thermal state in advance and pre-adjust the cooling equipment's operating state before the hot spot temperature approaches the control threshold, achieving predictive drive control. This strategy effectively reduces control lag, improves the accuracy of hot spot temperature control, and reduces frequent start-stop phenomena of the cooling equipment, achieving a coordinated optimization of control performance and operational economy.

[0044] In one embodiment, to ensure control quantity To achieve optimal results, this embodiment optimizes the process by constructing an optimization function. ; In the formula: To control performance indicators; and These are the weighting coefficients for the hotspot temperature control error and the control quantity, respectively. Determine the control quantity according to the constructed optimization function F. .

[0045] In one embodiment, the execution module needs to evaluate the cooling device before driving it. In this embodiment, the evaluation method for the cooling device includes: In this embodiment, the cumulative operating time of the cooling equipment is used as the main evaluation index. Under the premise of meeting the cooling capacity requirements, the equipment with the shortest cumulative operating time is given priority to achieve a balanced distribution of the operating time of each cooling unit. When obtaining the cumulative operating time of the cooling equipment, the following method is used to calculate the [number of]th [times]. The cumulative operating time of the cooling equipment is: ; in, This is a function for the device's operating status when the device is put into operation. Not put into operation .

[0046] After obtaining the cumulative operating time of all cooling equipment At that time, priority is defined based on the cumulative running time, i.e., priority sorting; the cooling equipment is started according to the priority sorting to cool down the transformer in advance.

[0047] In one embodiment, when the cooling equipment needs to be activated, all operable equipment sets... The cooling equipment (selected according to priority) is sorted by cumulative operating time, and the equipment with the shortest cumulative operating time is selected for operation first. ; in, This indicates the number of the cooling equipment that will be put into operation first.

[0048] Since there is a possibility of failure during the operation of cooling equipment, this embodiment sets fault constraints in the control strategy to ensure the reliability of system operation. When the operating equipment fails or receives a protection action signal, it immediately exits operation and automatically switches to the standby equipment. The control logic is expressed as follows: in, This indicates a equipment malfunction. Number the backup equipment.

[0049] Where: Fault i Fault status of the i-th cooler; i =1: indicates that the i-th cooler has malfunctioned; u i : The operating status of the i-th cooler, u i =1 indicates that it is in operation, u i =0 indicates exiting the program; uj Ω: Operating status of standby cooler j; Ω: Set of standby coolers; So: ; This means that when the i-th cooler fails, the failed cooler should be taken out of service immediately.

[0050] and: ; This indicates that a backup cooler is being added from the backup cooler set Ω.

[0051] Therefore, from a control logic perspective, it is: fault detection → faulty equipment shutdown → backup equipment activation.

[0052] To evaluate the balancing effect of the rotation strategy, the equipment operation balancing index is defined as: ; in, This represents the total number of cooling devices. For the first Cumulative operating time of the cooling equipment; This represents the average cumulative operating time of all cooling equipment.

[0053] This evaluation index reflects the dispersion of the operating time of each piece of equipment. The smaller the value, the more evenly the equipment operates, and the better the rotation strategy works; therefore, the optimization objective of intelligent rotation control can be expressed as:

[0054] That is, under the conditions of meeting cooling requirements and equipment operation constraints, the cumulative operating time of each cooling device should be as balanced as possible, thereby reducing excessive wear of individual devices and improving the overall operational reliability and service life of the system.

[0055] In this embodiment, to improve the operational reliability and fault self-recovery capability of the cooling control system, a digital fault diagnosis and fault-tolerant control strategy is constructed. By collecting key operating parameters such as temperature, current and communication status in real time, the health status of the equipment is comprehensively judged, and corresponding fault-tolerant control measures are executed based on the diagnosis results to achieve automatic isolation of faulty equipment and rapid deployment of backup equipment, ensuring the continuous and stable operation of the cooling system.

[0056] The device status determination function is ; in: This refers to the equipment's temperature status. This refers to the operating current status of the equipment. This indicates the communication status.

[0057] Based on monitoring data and preset thresholds, the operating status of the equipment can be classified as follows: ; The higher the status value, the higher the degree of equipment abnormality.

[0058] When any monitored parameter exceeds the set threshold or communication is interrupted, the system immediately determines that the device has entered an abnormal state. The criterion can be expressed as follows: ; in: Temperature alarm threshold; The current alarm threshold; This indicates a communication error or interruption. The symbol 'v' represents the logical 'OR'. The three conditions represent: ； This indicates that the temperature has exceeded the permissible alarm threshold: T represents the currently monitored device temperature; This is the temperature alarm threshold. As long as: This indicates that the equipment temperature is abnormal. This indicates that the equipment's operating current exceeds the permissible alarm threshold: I represents the current operating current of the equipment; This indicates the current alarm threshold.

[0059] For example, when the cooler fan or oil pump motor experiences overload or mechanical jamming, the operating current may rise abnormally. This condition can be used to identify the corresponding fault. C=0: indicates communication abnormality or interruption; C=1: communication normal; C=0: communication abnormality or interruption. Therefore, even if the temperature and current do not exceed the threshold, the equipment will be judged as abnormal as long as communication is interrupted.

[0060] When equipment is determined to be in a faulty state, the control system immediately stops the operation of the faulty equipment and automatically switches to the backup equipment. Its fault-tolerant control logic is expressed as follows: ; in: The operating status of the faulty equipment; This is the standby equipment operating state. When the state S of equipment i is determined to be a fault state (S=2), the control system immediately sets the operating state of the i-th cooling equipment to u. i =0, meaning stop operation; simultaneously, set the operating status of standby device j to u. j=1, meaning it is put into operation, thereby realizing the automatic switching between the shutdown of faulty equipment and the activation of standby equipment.

[0061] To evaluate system operational reliability, the fault recovery success rate is defined as: ; in: This represents the total number of system failures. This is the number of times the system has successfully completed automatic recovery.

[0062] A higher recovery success rate indicates that the digital fault diagnosis and fault-tolerant control strategy is more effective, and the system can quickly switch equipment after a fault occurs, reducing the cooling capacity interruption time and improving the safety, reliability and continuous operation capability of the transformer cooling system.

[0063] In one embodiment, after obtaining the cumulative operating time of the cooling equipment, a health evaluation of the cooling equipment is required. During the evaluation process, a health evaluation function is established, and the evaluation is performed using the health evaluation function. In this embodiment, the health evaluation function is: ; In the formula: for Fault judgment values ​​for each device; For the first The temperature of each device. For the first The operating current of each device. For the first Communication status of each device , , These are the weighting coefficients.

[0064] when When the equipment health assessment value exceeds the threshold When a fault occurs, the system automatically determines that the equipment is faulty, immediately disconnects the faulty equipment, removes it from the cooling equipment, and then re-prioritizes it based on the accumulated time.

[0065] As can be seen from the above embodiments, the present invention solves the following problems associated with traditional transformer cooling systems that commonly employ start-stop control strategies based on fixed temperature thresholds: 1) Response lag: Due to a 20-40 minute time delay between oil temperature changes and load changes, the cooling system cannot respond promptly to sudden changes in heat load; 2) Low energy efficiency: The system design leaves a large margin to ensure reliability under extreme conditions, resulting in a "powered engine for a small load" situation during most normal operating periods; 3) Overcooling: Traditional hysteresis loop control cannot establish a precise thermal balance relationship, and the cooling capacity provided by the cooling system often exceeds actual demand, resulting in significant energy waste; 4) Lack of predictive capability: The cooling strategy cannot be adjusted in advance according to load trends, leading to drastic temperature fluctuations under dynamic conditions such as peak shaving; 5) Lack of complete independent control over the core technology of the controller. Example 2

[0066] This embodiment provides a digital cooling control method for transformers that integrates hotspot prediction, including: Real-time acquisition of multi-source status data of the transformer, including at least top oil temperature, ambient temperature and load current; The predicted hot spot temperature of the transformer is obtained based on the multi-source state data collected at all sampling times. The hot spot temperature prediction value is used to determine whether there is an over-temperature risk; when it is determined that there is an over-temperature risk, the obtained hot spot temperature prediction value is used as the feedforward input of the PID controller to generate a cooling control command. The cooling equipment is driven to operate according to the cooling control commands generated by the control module. When driving the cooling equipment, the cooling equipment to be put into operation is determined according to the priority of the cumulative running time of the cooling equipment from shortest to longest.

[0067] When a cooling device malfunctions, all backup cooling devices are retrieved, and the cumulative operating time of all backup cooling devices is used to determine whether the malfunctioning cooling device should be replaced.

[0068] Methods for determining whether there is an overheating risk based on hotspot temperature predictions include: calculating the difference between the hotspot temperature prediction and the target hotspot value to obtain the degree of deviation, and judging whether an overheating risk has occurred based on the obtained degree of deviation.

[0069] Methods for generating cooling control commands include: The difference between the predicted hot spot temperature and the target hot spot temperature is used as the control error, and the control error and its rate of change are used as the input of the PID controller. The PID controller corrects its proportional, integral, and derivative coefficients online according to preset fuzzy inference rules. The PID controller outputs a control quantity based on the corrected proportional coefficient, integral coefficient, and derivative coefficient, and the control quantity constitutes the cooling control command. Example 3

[0070] like Figure 7 The cooler integrated control unit 10 shown in this embodiment includes an MCU board 1, a submersible pump control board 2, a fan control board 3, a fan control board 4, a fan control board 5, a fan control board 6, a power control board I 7, a power control board II 8, and an I / O control board 9. Each control board is combined into a 4U chassis via a backplane bus using PCB plug-in terminals.

[0071] MCU board 1 is connected to the adjacent integrated control unit of the cooler via a communication bus. It is responsible for the real-time operation of the predictive drive control algorithm, executing hot spot temperature prediction and control decisions, managing communication between adjacent units and data interaction at the station control layer, evaluating cooling requirements, issuing intelligent scheduling and fault tolerance information, and realizing data storage.

[0072] The submersible pump control board 2 is connected to the submersible pump motor on the transformer cooler. It is responsible for driving and protecting the submersible pump motor. The drive adopts a parallel circuit of bidirectional thyristors and bypass switches. The domestic MCU on the control board realizes zero-crossing detection, bidirectional thyristor triggering, and bypass switch control logic. The protection is realized by collecting information from the current transformer and voltage transformer to realize the motor overload, phase loss, short circuit, grounding, and stall protection functions. It also receives motor adjustment commands issued by the MCU board 1.

[0073] Fan control boards 3, 4, 5 and 6 are respectively connected to the corresponding fan motors on the transformer cooler, and are responsible for driving and protecting the fan motors. Their implementation and functions are the same as those of submersible pump control board 2.

[0074] I. Power control board 7 is connected to the 400V three-phase power output side of the #1 plant service transformer (station service transformer) of the power plant (substation). The control board MCU collects information from the current transformer and voltage transformer to realize the monitoring and metering of the three-phase input power and shares data in real time through the backplane communication bus. If a power fault is detected, the power is cut off and the power fault information of this segment is issued. II. Power control board 8 is put into operation.

[0075] II. Power control board 8 is connected to the 400V three-phase power output side of the #2 plant service transformer (station service transformer) of the power plant (substation). The control board MCU collects information from the current transformer and voltage transformer to realize the monitoring and metering of the three-phase input power and shares data in real time through the backplane communication bus. If a power fault is detected, the power is cut off and the power fault information of this segment is issued. I. Power control board 7 is put into operation.

[0076] The I / O control board 9 connects to the transformer body's top-level oil temperature sensor, ambient temperature sensor, bushing CT secondary signal, cooler oil pump oil flow relay signal, and high-voltage circuit breaker position status information, completing data preprocessing mapping consisting of multi-source monitoring data acquisition, outlier removal, filtering, and status fusion. Cooler fan oil pump alarm information is output as a signal connected to the backend monitoring computer. Example 4

[0077] like Figure 8 The embodiment shown is a digital control cabinet for transformers that integrates hotspot prediction, consisting of four integrated cooling control units and one display unit. It controls four sets of coolers on a large power transformer and achieves temperature control during transformer operation through hotspot prediction.

[0078] The motor (oil pump / fan) control board output of the #1 cooler integrated control unit 11 is connected to one oil pump and four fans on the #1 cooler of the transformer. The first communication port is connected to the adjacent cooler integrated control unit, the second communication port is connected to the display unit 15, and the third communication port is connected to the background monitoring through the switch.

[0079] The motor (oil pump / fan) control board output of the #2 cooler integrated control unit 12 connects to one oil pump and four fans on the #2 cooler via the transformer, and the communication connection is the same as that of the #1 cooler integrated control unit 11.

[0080] The motor (oil pump / fan) control board output of the #3 cooler integrated control unit 13 connects to one oil pump and four fans on the #3 cooler via the transformer, and the communication connection is the same as that of the #1 cooler integrated control unit 11.

[0081] The motor (oil pump / fan) control board output of the #4 cooler integrated control unit 14 connects to one oil pump and four fans on the #4 cooler via the transformer, and the communication connection is the same as that of the #1 cooler integrated control unit 11.

[0082] Each cooler integrated control unit adopts a masterless communication and collaboration method to achieve data sharing and decentralized control.

[0083] Engineering verification and performance analysis 1) Engineering Application Platform To verify the effectiveness of the proposed predictive-driven digital cooling control method, an engineering application platform was built based on the cooling system of the main transformer in a 750 kV substation. The platform uses a domestically produced industrial-grade MCU as the control core, integrating functions such as temperature acquisition, load monitoring, fan and oil pump control, communication management, and host computer monitoring. It achieves real-time acquisition of multi-source operating data, edge computing, and control execution. The actual scenario is shown in [link to physical scenario]. Figure 9 As shown.

[0084] During the experiment, the system collected real-time data on top-layer oil temperature, ambient temperature, load current, and the operating status of the cooling equipment. Based on the predictive-driven digital cooling control method proposed in this invention, it completed hotspot temperature prediction, cooling demand assessment, intelligent equipment scheduling, and fault-tolerant control. To objectively evaluate the control performance, a traditional fixed-threshold control strategy was used as a comparison, and comprehensive verification was conducted from aspects such as hotspot temperature prediction accuracy, control dynamic performance, operational economy, and system reliability.

[0085] 2) Validation of hotspot temperature prediction model To verify the accuracy of the hotspot temperature prediction model, field operating data under different load conditions were selected for testing, and the prediction results were compared and analyzed with online monitoring data. Prediction performance was evaluated using mean absolute error (MAE), root mean square error (RMSE), and mean absolute percentage error (MAPE), calculated as follows: ; ; ; In the formula, Let m be the measured hotspot temperature of the m-th sample. Let M be the predicted hotspot temperature of the m-th sample, where M is the number of samples.

[0086] like Figure 10 The figure shows a comparison curve between the predicted and measured hotspot temperatures. The results show that the prediction model can track the hotspot temperature change trend well, maintaining high prediction accuracy even during periods of rapid load change. This is because the IEC thermal model comprehensively considers factors such as load factor, top oil temperature, and cooling status, and can accurately describe the thermal inertia characteristics of the transformer. Therefore, it still has good predictive ability during periods of rapid load change. Statistical results are shown in Table 1.

[0087] Table 1 Hotspot Temperature Prediction Performance

[0088] Analysis shows that the hot spot temperature prediction method based on the IEC thermal model fully considers factors such as ambient temperature, load rate and cooling status, and can accurately reflect the changes in the thermal state of the transformer, providing a reliable basis for predictive drive control.

[0089] 3) Control performance and comprehensive evaluation To verify the overall performance of the predictive-driven digital cooling control method, comparative experiments were conducted using traditional fixed threshold control and the method presented in this paper under the same load conditions.

[0090] like Figure 11The diagram shows the hot spot temperature change curves under two control strategies. Traditional control relies on real-time feedback, and the cooling equipment is only activated after the hot spot temperature reaches the control threshold, resulting in a certain response lag. The method of this invention assesses the cooling demand in advance based on the hot spot temperature prediction results, enabling the fan and oil pump to be activated on demand, making the hot spot temperature change more stable and reducing temperature overshoot and fluctuation amplitude.

[0091] Meanwhile, to verify the intelligent scheduling strategy, the cumulative operating time of each fan and oil pump within a continuous operating cycle was statistically analyzed, such as... Figure 12 The experimental results show that the traditional rotation method has the problem of uneven distribution of equipment running time, while the intelligent scheduling strategy proposed in this invention can dynamically adjust the order of equipment input based on the cumulative running time, so that the running time of each piece of equipment is more balanced, which is conducive to reducing mechanical wear and extending the service life of equipment.

[0092] Further simulations of typical operating conditions such as fan failure, oil pump failure, and communication anomalies were conducted to verify the digital fault diagnosis and fault-tolerant control. Figure 13 The experimental results show that the system can quickly complete fault identification, faulty equipment isolation and backup equipment switching. The hot spot temperature remains stable throughout the fault handling process, which verifies the good continuous operation capability and reliability of the proposed method.

[0093] To comprehensively evaluate the control effect, indicators such as hot spot temperature prediction accuracy, control response time, hot spot temperature fluctuation range, auxiliary cooling energy consumption, equipment running time standard deviation, and fault switching time were selected for statistical analysis. The results are shown in Table 2.

[0094] Table 2 Overall Performance Comparison

[0095] Comprehensive analysis shows that the predictive-driven digital cooling control method proposed in this invention fully leverages the synergistic effects of hotspot temperature prediction, predictive control, intelligent scheduling, and fault tolerance. Compared with traditional fixed threshold control, the proposed method improves the accuracy of hotspot temperature prediction, enhances the system's dynamic response capability, reduces hotspot temperature fluctuations and auxiliary cooling energy consumption, and achieves balanced operation of cooling equipment and rapid fault switching, thereby comprehensively improving the economic efficiency, reliability, and intelligence level of the cooling system. Engineering verification results demonstrate that this method has significant engineering application value and can provide effective technical support for the digital cooling control of large power transformers.

[0096] Example embodiments have been disclosed herein, and while specific terminology has been used, it is for illustrative purposes only and should be construed as such, and is not intended to be limiting. In some instances, it will be apparent to those skilled in the art that features, characteristics, and / or elements described in conjunction with particular embodiments may be used alone, or in combination with features, characteristics, and / or elements described in conjunction with other embodiments, unless otherwise expressly indicated. Therefore, those skilled in the art will understand that various changes in form and detail may be made without departing from the scope of the invention as set forth in the appended claims.

Claims

1. A digital cooling control system for transformers integrating hotspot prediction, characterized in that, include: The sensing module is used to collect multi-source status data of the transformer in real time. The multi-source status data includes at least the top oil temperature, ambient temperature and load current. The prediction module obtains the predicted hot spot temperature of the transformer based on the multi-source state data collected at all sampling times; The control module determines whether there is an over-temperature risk based on the hot spot temperature prediction value; when it determines that there is an over-temperature risk, it uses the obtained hot spot temperature prediction value as the feedforward input of the PID controller to generate a cooling control command. The execution module drives the cooling equipment to operate according to the cooling control commands generated by the control module. When driving the cooling equipment, the cooling equipment to be put into operation is determined according to the priority of the cumulative running time of the cooling equipment from shortest to longest.

2. The transformer digital cooling control system with integrated hotspot prediction as described in claim 1, characterized in that, The prediction module uses the following methods to obtain the predicted hot spot temperature of the transformer: The load factor is obtained from the ratio of load current to rated current in the multi-source status data. The top oil temperature rise is obtained based on the load rate and the rated top oil temperature rise. The winding hot spot temperature rise is obtained based on the load rate and the rated hot spot temperature rise. Based on the obtained top oil temperature rise, winding hot spot temperature rise, and ambient temperature, the predicted value of the hot spot temperature is obtained.

3. The transformer digital cooling control system with integrated hotspot prediction as described in claim 1, characterized in that, In the control module, the method for generating cooling control commands by using the obtained hot spot temperature prediction value as the feedforward input of the PID controller includes: The difference between the predicted hot spot temperature and the preset target hot spot temperature is used as the hot spot temperature control error, and the hot spot temperature control error and its rate of change are used as the input of the PID controller. The PID controller corrects its proportional, integral, and derivative coefficients online according to preset fuzzy inference rules. The PID controller outputs a control quantity based on the corrected proportional coefficient, integral coefficient, and derivative coefficient, and the control quantity constitutes the cooling control command.

4. The transformer digital cooling control system with integrated hotspot prediction according to claim 3, characterized in that, The control outputs of the PID controller include: ； In the formula: To control the quantity; This is due to the temperature control error of the hot spot. This is the proportionality coefficient; The integral coefficient; These are the differential coefficients; The hot spot temperature control error It is obtained by pre-setting target values ​​and predicted values ​​of hotspot temperatures.

5. The transformer digital cooling control system with integrated hotspot prediction according to claim 4, characterized in that, The control quantity Determined in the following manner; Using control variables and hot spot temperature control error Construct an optimization function F; the optimization function F is: ； In the formula: To control performance indicators; and These are the weighting coefficients for the hotspot temperature control error and the control quantity, respectively. Determine the control quantity according to the constructed optimization function F. .

6. The transformer digital cooling control system with integrated hotspot prediction according to claim 1, characterized in that, The execution module further includes the following before driving the cooling device: A health evaluation function for each cooling device is constructed based on its device temperature, operating current, and communication status. The constructed health evaluation function is used to evaluate the cooling equipment and select all target cooling equipment. Prioritize all target cooling devices by using the cumulative operating time of each target cooling device.

7. A digital cooling control method for transformers integrating hotspot prediction, characterized in that, include: Real-time acquisition of multi-source status data of the transformer, including at least top oil temperature, ambient temperature and load current; The predicted hot spot temperature of the transformer is obtained based on the multi-source state data collected at all sampling times. Based on the hot spot temperature prediction value, it is determined whether there is an over-temperature risk; when it is determined that there is an over-temperature risk, the obtained hot spot temperature prediction value is used as the feedforward input of the PID controller to generate a cooling control command. The cooling equipment is driven to operate according to the cooling control commands generated by the control module. When driving the cooling equipment, the cooling equipment to be put into operation is determined according to the priority of the cumulative running time of the cooling equipment from shortest to longest.

8. The digital cooling control method for transformers based on hotspot prediction according to claim 7, characterized in that, When a cooling device malfunctions, all backup cooling devices are retrieved, and the cumulative operating time of all backup cooling devices is used to determine whether the malfunctioning cooling device should be replaced.

9. The digital cooling control method for transformers based on hotspot prediction according to claim 7, characterized in that, Methods for determining whether there is an overheating risk based on hotspot temperature predictions include: calculating the difference between the hotspot temperature prediction and the target hotspot temperature to obtain the degree of deviation, and judging whether an overheating risk has occurred based on the obtained degree of deviation.

10. The digital cooling control method for transformers based on hotspot prediction according to claim 7, characterized in that, Methods for generating cooling control commands include: The difference between the predicted hot spot temperature and the target hot spot temperature is used as the control error, and the control error and its rate of change are used as the input of the PID controller. The PID controller corrects its proportional, integral, and derivative coefficients online according to preset fuzzy inference rules. The PID controller outputs a control quantity based on the corrected proportional coefficient, integral coefficient, and derivative coefficient, and the control quantity constitutes the cooling control command.

Images