AC short-circuit ice-melting control method and ice-melting distribution transformer

Abstract

The present invention relates to the technical field of power grid deicing, and in particular to an AC short-circuit ice-melting control method and an ice-melting distribution transformer. The present invention includes a distribution transformer with dual functions of voltage reduction and ice-melting, an ice-melting control system and a control method thereof. The transformer is provided with two sets of low-voltage output terminals, which are used for power distribution and ice-melting operations respectively; the ice-melting control system includes a local intelligent control box and a remote communication control box, and uses a deep reinforcement learning model to calculate the optimal ice-melting power level and ice-melting time; the control method obtains environmental parameters and equipment parameters, and combines the coordinated control of boundary switches, ice-melting switches, protection switches and short-circuit switches to achieve precise regulation of the ice-melting process. The present invention realizes a one-button ice-melting function, and has the characteristics of simple operation, high ice-melting efficiency, and remote control. It significantly improves the intelligence level and safety and reliability of the power grid ice-melting operation, and provides an innovative solution for the anti-icing management of the power system.

Description

Technical Field

[0001] The invention relates to the technical field of power grid deicing, in particular to an AC short-circuit deicing control method and a deicing distribution transformer. Background Art

[0002] The ice accumulation on distribution lines poses a major challenge to the stable operation of the power system, especially under severe climatic conditions. At present, anti-icing and de-icing measures for distribution lines mainly rely on traditional manual observation and manual ice knocking operations. Although this method can alleviate the threat of ice accumulation to a certain extent, its inherent limitations are obvious. In cold areas or extreme weather conditions, the ice layer formed on the surface of power lines not only increases the line load, but may also cause serious accidents such as line breakage and pole collapse, threatening the safe and stable operation of the power grid.

[0003] At present, the de-icing methods for distribution network lines mainly include manual ice knocking, thermal ice melting and mechanical de-icing. Among them, although the manual ice knocking method is intuitive and simple, it has problems such as manual ice observation and ice knocking, which are not only inefficient, but also difficult to fully cover the wide distribution network line network, low operating efficiency, high safety risks, and limited coverage. Especially under severe weather conditions, a large amount of manpower investment not only increases the cost of operation and maintenance, but also makes it difficult to ensure timely handling of ice accumulation hazards throughout the network. Since distribution network lines often span a vast area and are intricately distributed, a large amount of human resources and time costs are required to achieve real-time monitoring and rapid response to the ice accumulation of each line. This not only increases the difficulty of operation and maintenance work, but also limits the possibility of timely and effective handling of ice accumulation problems.

[0004] In order to meet this challenge, the industry has carried out a series of research and development work on ice-melting technologies, such as the "A distribution network line ice-melting method and system based on distribution transformers in substations" disclosed in patent application number CN119093258A. However, these solutions still face many problems in practical applications: First, the operation process of the existing ice-melting system is cumbersome, and it is necessary to manually open and close the line switches one by one, which is not only inefficient but also increases the operational risks; secondly, due to the lack of special modification of the distribution transformer, the ice-melting distance is limited, and the dual functions of distribution and ice-melting cannot be realized; thirdly, the lack of intelligent communication protocols and control algorithms leads to insufficient system reliability and adaptability.

[0005] In addition, existing ice-melting control systems generally have problems such as weak data collection and analysis capabilities, lack of remote monitoring functions, and low fault diagnosis efficiency. Under complex climatic conditions, these technical limitations seriously affect the effectiveness of ice-melting operations and the reliability of power grid operation. Summary of the invention

[0006] In view of the problems existing in the prior art, the present invention is proposed.

[0007] Therefore, the problem to be solved by the present invention is how to solve the problems commonly existing in the existing ice melting control system, such as weak data collection and analysis capabilities, lack of remote monitoring function, low fault diagnosis efficiency, etc.

[0008] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0009] In a first aspect, an embodiment of the present invention provides an AC short-circuit de-icing control method, which includes obtaining environmental parameters and equipment parameters of a de-icing line, wherein the environmental parameters include temperature, humidity, wind speed, and ice thickness, and the equipment parameters include rated capacity, power factor, and number of temperature sensors of a de-icing transformer;

[0010] Based on the environmental parameters and equipment parameters, a deep reinforcement learning model is used to calculate the optimal ice-melting power level and ice-melting time; the deep reinforcement learning model includes:

[0011] An Actor network, used for selecting an ice-melting power gear action according to a state, the Actor network comprising an input layer with the same dimension as the state, two hidden layers of 256 nodes, and an output layer corresponding to three ice-melting power gears;

[0012] A critic network, used to evaluate the rationality of the selected power level, the critic network comprising an input layer of state and action information, two hidden layers of 256 nodes, and an output layer representing the Q value;

[0013] The deep reinforcement learning model also includes: an exploration strategy based on visibility, whose exploration probability ε satisfies:

[0014] ,

[0015] in is the current visibility in hundred meters;

[0016] According to the optimal ice-melting power level and ice-melting time, the output voltage and switch sequence of the ice-melting transformer are controlled, wherein the switch sequence includes the action timing of the boundary switch, the ice-melting switch, the protection switch and the short-circuit switch;

[0017] Calculate the temperature rise sensitivity coefficient, which is calculated based on the load rate scenario division and temperature fluctuation of the ice melting transformer for 7 operating days;

[0018] The operation data of the ice-melting transformer for 7 operation days are divided into six load rate scenarios according to [0, 20%], (20%, 40%], (40%, 60%], (60%, 80%], (80%, 100%], and 100%. Based on the relationship between the power fluctuation value and the temperature fluctuation value, the temperature rise sensitivity coefficient under each scenario is calculated;

[0019] Based on the temperature rise sensitivity coefficient, the current value during the ice melting process is monitored in real time. When the current value reaches the preset ice melting current demand value, the current voltage level is maintained; when the preset ice melting current demand value is not reached, the voltage level is adjusted until the requirement is met.

[0020] As a preferred solution of the AC short-circuit de-icing control method of the present invention, the following is used: the deep reinforcement learning model is used to perform adaptive de-icing power control: the de-icing power is discretized into three actions: fast, medium and slow; the change in ice thickness is predicted according to environmental parameters; the line ice load and the risk of pole collapse are calculated; the de-icing power is controlled based on the balance between risk and cost, and the risk and cost balance uses a reward function to control the de-icing power, wherein the reward function of the risk and cost balance needs to use the coefficient value of the de-icing cost;

[0021] The prediction of ice thickness change adopts the following formula:

[0022] ,

[0023] in, is the predicted new ice thickness; is the current ice thickness; is the ice thickness coefficient; is the reference temperature; is the ambient temperature; is the relative humidity of the environment; is the reference humidity; is the wind speed; For time;

[0024] The line icing load is calculated using the following formula:

[0025] ,

[0026] Where ρ is the density of ice; D is the diameter of the transmission line; is the ice thickness; L is the length of the transmission line between adjacent towers;

[0027] The pole breaking risk P is calculated using the following formula:

[0028] ,

[0029] in, is the ice load; is the rated ice load;

[0030] The ice melting cost C is calculated using the following formula:

[0031] ,

[0032] in, For ice melting costs; is the function of power and cost; is the ice thickness;

[0033] The reward function that balances risk and cost Use the following formula:

[0034] ,

[0035] in, Used to balance the impact of risk and cost; and is the weight coefficient; To reduce the risk of pole breaking; For ice melting costs.

[0036] As a preferred solution of the AC short-circuit ice melting control method of the present invention, the control of the switch sequence includes:

[0037] Before the ice melting operation begins, isolate the non-ice melting area through the boundary switch;

[0038] According to the preset program, the start and adjustment of the ice melting current is controlled by the ice melting switch;

[0039] Monitor the line status in real time through the protection switch and cut off the power supply when an abnormality is detected;

[0040] The terminal line is short-circuited by the short-circuit switch to form an ice-melting circuit.

[0041] As a preferred solution of the AC short-circuit ice-melting control method of the present invention, it further includes: calculating and monitoring the ice-melting efficiency, wherein the ice-melting efficiency is the ratio of the actual ice-melting power to the maximum ice-melting power;

[0042] When the ice melting efficiency is lower than the preset threshold, the optimal ice melting power level is adjusted.

[0043] As a preferred solution of the AC short-circuit ice-melting control method of the present invention, it also includes remote control: establishing a communication link with a remote short-circuit communication control box through a local intelligent control box; and realizing remote monitoring, fault diagnosis and emergency response.

[0044] As a preferred solution of the AC short-circuit ice-melting control method described in the present invention, during the ice-melting process, when the ambient temperature is higher than a preset temperature threshold or the ice thickness is less than a preset thickness threshold, the ice-melting process is automatically stopped, and the ice-melting time and power data are recorded.

[0045] In a second aspect, an embodiment of the present invention provides an ice-melting distribution transformer, which includes a transformer body, wherein the transformer body is configured with two groups of low-voltage output terminals, a first group of normal voltage output terminals designed for power distribution users, and a second group of specific voltage output terminals designed for ice-melting operations;

[0046] The two groups of low-voltage output terminals have two states, one is to work alone, and the other is to melt ice and supply power to low-voltage users at the same time; the output voltages of the two groups of output terminals meet the following requirements:

[0047] ,

[0048] in, is the voltage at the output end (normal voltage for power distribution users or voltage for ice melting); is the number of coil turns at the output end; is the number of coil turns at the input end; is the voltage at the input terminal;

[0049] The ice melting power at the specific voltage output end of the ice melting operation design satisfies:

[0050] ,

[0051] in, is the ice melting power; It is the voltage at the output terminal of the voltage used for ice melting; It is the current when the ice melts;

[0052] An ice-melting control box, including a local intelligent control box and a remote short-circuit communication control box, wherein the local intelligent control box is provided with a one-touch ice-melting button;

[0053] The ice-melting power regulating device is provided with a plurality of gears of ice-melting power output at the ice-melting output end; the efficiency of the ice-melting power output satisfies:

[0054] ,

[0055] in, is the ice melting efficiency; is the actual ice melting power used; is the ice melting power when the transformer is in the highest gear;

[0056] The switch control system includes a boundary switch, an ice-melting switch, a protection switch and a short-circuit switch.

[0057] As a preferred solution of the ice-melting distribution transformer described in the present invention, the remote short-circuit communication control box is connected to the local intelligent control box through a high-speed communication link to achieve remote synchronous control.

[0058] As a preferred solution of the ice-melting distribution transformer described in the present invention, the local intelligent control box is equipped with a deep reinforcement learning model operation unit for calculating the optimal ice-melting power level and ice-melting time.

[0059] As a preferred solution of the ice-melting distribution transformer described in the present invention, the remote short-circuit communication control box is connected to the local intelligent control box through a high-speed communication link to achieve remote synchronous control.

[0060] The beneficial effects of the present invention are as follows: the present invention provides an AC short-circuit ice-melting control method and an ice-melting distribution transformer, which realizes a perfect combination of power supply and ice-melting functions by setting a structural design of dual output ends of the distribution transformer, solves the problem that the traditional solution requires a special ice-melting transformer or temporary modification, and achieves the effect of one machine with two uses and saving investment.

[0061] By introducing an ice-melting control strategy based on deep reinforcement learning, the system can autonomously learn the optimal ice-melting power level and time, overcoming the defect that traditional fixed strategies cannot adapt to complex climatic conditions, realizing the intelligent and precise control of the ice-melting process, and significantly improving the ice-melting efficiency.

[0062] By designing a multi-level optimal ice-melting power gear adjustment mechanism and combining it with real-time calculation of the temperature rise sensitivity coefficient, the system can flexibly adjust the power output according to the actual icing conditions, avoiding the energy waste or insufficient ice melting caused by rough power regulation in traditional solutions, and achieving the effect of refined control, energy saving and environmental protection.

[0063] By incorporating the coordinated control strategy of boundary switches, ice-melting switches, protection switches and short-circuit switches, the system realizes fully automated operation of the ice-melting process, eliminating the safety hazards brought about by traditional manual operation, while significantly improving operating efficiency and realizing safe, reliable, efficient and convenient ice-melting operations.

[0064] By establishing a two-way real-time communication mechanism between the local intelligent control box and the remote communication control box, the system has the ability of remote monitoring, fault diagnosis and emergency handling, overcoming the problem of insufficient remote management capabilities in traditional solutions and realizing intelligent operation and maintenance and rapid response.

[0065] By designing a real-time monitoring and automatic adjustment mechanism for ice melting efficiency, the system can dynamically optimize ice melting parameters, ensure the quality of ice melting while avoiding energy waste, and achieve the unity of economy and efficiency in the ice melting process.

[0066] It realizes the one-key ice-melting function, simplifies the operation process, improves the ice-melting efficiency, and the ice-melting time is only 1 hour, which can meet a wider range of ice-melting needs. Compared with the traditional manual knocking method, the efficiency is greatly improved and the safety risk is reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0068] Figure 1 A flow chart of an AC short-circuit de-icing control method and a de-icing distribution transformer;

[0069] Figure 2 Computer equipment diagram for AC short circuit de-icing control method and de-icing distribution transformer;

[0070] Figure 3 It is a schematic diagram of an AC short-circuit ice-melting control method and an ice-melting distribution transformer;

[0071] Figure 4 It is a schematic diagram of a local controller for an AC short-circuit ice-melting control method and an ice-melting distribution transformer;

[0072] Figure 5 The present invention is a schematic diagram of a remote controller for an AC short-circuit de-icing control method and a de-icing distribution transformer. DETAILED DESCRIPTION

[0073] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0074] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0075] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor is it a separate or selective embodiment that is mutually exclusive with other embodiments.

[0076] Example 1 Reference Figure 1 and Figure 2 , which is the first embodiment of the present invention, provides an AC short-circuit ice-melting control method and an ice-melting distribution transformer, including:

[0077] S100: Acquire environmental parameters and equipment parameters of the ice-melting line. The environmental parameters include temperature, humidity, wind speed, and ice thickness. The equipment parameters include rated capacity, power factor, and number of temperature sensors of the ice-melting transformer.

[0078] In the embodiment of the present application, the environmental parameters are acquired by multi-source data acquisition. The temperature uses a PT100 temperature sensor with a measurement range of -50℃~150℃ and an accuracy of ±0.1℃; the humidity uses a capacitive humidity sensor with a measurement range of 0%~100%RH and an accuracy of ±2%RH; the wind speed uses an ultrasonic wind speed sensor with a measurement range of 0~60m / s and an accuracy of ±0.1m / s; the ice thickness uses a laser ranging sensor with a measurement range of 0~500mm and an accuracy of ±1mm.

[0079] In an optional embodiment, the installation positions of the temperature sensors can be divided into three categories: line temperature sensor, ambient temperature sensor and transformer temperature sensor. The line temperature sensor is installed on the surface of the conductor to directly monitor the conductor temperature; the ambient temperature sensor is installed on the tower, 2 meters above the ground, to monitor the ambient temperature; the transformer temperature sensor is installed on the transformer body to monitor the transformer operating temperature.

[0080] Specifically, the acquisition of equipment parameters is mainly achieved through the distribution automation system. The rated capacity of the ice-melting transformer is read from the nameplate information, usually 100kVA, 200kVA or 400kVA; the power factor is measured in real time by the power quality analyzer, generally required to be no less than 0.95; the number of temperature sensors is determined according to the actual installation data, and at least one temperature sensor is configured at each monitoring point.

[0081] It should be noted that the collection of environmental parameters and equipment parameters adopts real-time collection mode with a sampling period of 1 minute. All parameter data are pre-processed by the on-site controller and uploaded to the cloud data center in real time through the 4G communication network. The system verifies the collected data in real time. When abnormal data is found, the data compensation mechanism is automatically activated to ensure the continuity and reliability of the data.

[0082] In the embodiment of the present application, the field controller uses an STM32F407 processor with a main frequency of 168MHz, built-in 512KBFlash and 192KBRAM. The controller is equipped with a 7-inch touch screen display interface, IP65 protection level, and an operating temperature range of -40℃~85℃. The controller has the following functional modules:

[0083] Data acquisition module: responsible for the collection of environmental parameters and equipment parameters, supporting analog and digital input.

[0084] Data processing module: pre-process the collected data, including filtering, verification and abnormality judgment.

[0085] Communication module: supports RS485, Ethernet and 4G communications to achieve remote data transmission.

[0086] Storage module: uses 32GB industrial-grade SD card, which can store 3 months of historical data.

[0087] Human-machine interface module: provides parameter display, alarm prompt and manual control functions.

[0088] S200: Based on environmental parameters and equipment parameters, a deep reinforcement learning model is used to calculate the optimal ice-melting power level and ice-melting time;

[0089] S201: Adopt deep reinforcement learning model for adaptive control of ice melting power: discretize ice melting power into three actions: fast, medium and slow; predict ice thickness changes based on environmental parameters; calculate line ice load and pole failure risk; control ice melting power based on risk and cost balance.

[0090] In the embodiment of the present application, the deep reinforcement learning model is implemented using the DDPG (Deep Deterministic Policy Gradient) algorithm. The algorithm consists of two parts: the Actor network and the Critic network: the Actor network is responsible for selecting actions based on the state, that is, determining the optimal ice melting power level based on environmental parameters and equipment parameters; the Critic network is responsible for evaluating the value of the action, that is, evaluating the rationality of the selected power level.

[0091] Specifically, the Actor network adopts a three-layer fully connected network structure. The number of nodes in the input layer is the same as the state dimension, including parameters such as temperature, humidity, wind speed, and ice thickness; the hidden layer adopts a two-layer structure, with 256 nodes in each layer, and the activation function is ReLU; the number of nodes in the output layer is 3, corresponding to the three optimal ice melting power levels of fast, medium, and slow, and the activation function is .

[0092] The Critic network also uses a three-layer fully connected structure. The input layer contains state and action information; the hidden layer uses a two-layer structure, with 256 nodes in each layer and the activation function is ReLU; the output layer is a single node, representing the Q value. The network training uses the Adam optimizer, with learning rates of Actor1e-4 and Critic1e-3 respectively.

[0093] In an optional embodiment, in order to improve the exploration efficiency of the model, an exploration strategy based on visibility is introduced. When visibility is low, the exploration probability is increased so that the model tends to try different optimal ice melting power gears; when visibility is good, the exploration probability is reduced to make more use of the learned experience. The calculation formula of the exploration probability ε is:

[0094] ,

[0095] in is the current visibility in 100 meters. This dynamically adjusted exploration strategy significantly improves the model's adaptability under complex weather conditions.

[0096] It should be noted that the reward function design of the model takes into account multiple aspects:

[0097] Ice removal effect: positive rewards are given according to the degree of reduction in ice thickness; energy consumption: negative rewards are given according to the power consumption of the ice melting process; time consumption: negative rewards are given according to the duration of ice melting; safety constraints: large negative rewards are given when safety risks such as overheating occur.

[0098] S300: Control the output voltage and switch sequence of the ice-melting transformer according to the optimal ice-melting power level and ice-melting time. The switch sequence includes the action sequence of the boundary switch, the ice-melting switch, the protection switch and the short-circuit switch.

[0099] S301: Control of the switch sequence includes:

[0100] Before the ice melting operation begins, isolate the non-ice melting area through the boundary switch;

[0101] According to the preset program, the start and adjustment of the ice melting current is controlled by the ice melting switch;

[0102] Monitor the line status in real time through the protection switch and cut off the power supply when an abnormality is detected;

[0103] The terminal line is short-circuited by the short-circuit switch to form an ice-melting circuit.

[0104] In the embodiment of the present application, the control of the ice melting transformer adopts a hierarchical control architecture. The upper layer is the power gear control, which is responsible for determining the target power gear according to the output result of the deep reinforcement learning model; the middle layer is the voltage control, which is responsible for converting the power gear into a specific voltage setting value; the lower layer is the switch control, which is responsible for executing a specific switch action sequence.

[0105] Specifically, the power gear control includes three gears: fast, medium and slow:

[0106] Fast gear: corresponds to 90% to 100% rated power, suitable for severe icing conditions;

[0107] Medium speed: corresponds to 60% to 80% of rated power, suitable for medium icing conditions;

[0108] Slow gear: corresponds to 30% to 50% of rated power, suitable for light icing conditions.

[0109] In an optional embodiment, the voltage control adopts a fuzzy PID control strategy. The controller input is the deviation between the target power and the actual power. and the rate of change of deviation , the output is the voltage adjustment, and the control rules are as follows:

[0110] When | |When it is larger, increase the proportional gain of the controller and speed up the response;

[0111] When | |When it is smaller, reduce the proportional gain and improve the control accuracy;

[0112] When | |When it is larger, increase the differential time to suppress overshoot;

[0113] When | |When it is small, increase the integration time to eliminate the steady-state error.

[0114] It should be noted that the switch control adopts a state machine design, including the following states:

[0115] Initial state: all switches are in the disconnected state; ready state: check the status signals of each switch; isolation state: the boundary switch operates to isolate the non-ice-melting area; ice-melting state: the ice-melting switch and the short-circuit switch operate in sequence; protection state: the protection switch operates when an abnormality occurs; reset state: all switches reset after ice-melting is completed; end state: the entire action sequence is completed.

[0116] In the embodiment of the present application, the switch action timing control is implemented using the real-time operating system FreeRTOS, which has the following characteristics:

[0117] Priority management: The protection switch task has the highest priority to ensure system safety; Task synchronization: Use the semaphore mechanism to ensure the accurate order of switch actions; Real-time guarantee: The response time of key tasks does not exceed 10ms; Fault handling: Task anomalies can be detected and handled in a timely manner.

[0118] S400: Real-time monitoring of the current value during the ice-melting process. When the current value reaches a preset ice-melting current demand value, the current voltage level is maintained; if the current value does not reach the preset ice-melting current demand value, the voltage level is adjusted until the requirement is met.

[0119] S401: further comprising: calculating and monitoring ice melting efficiency, where the ice melting efficiency is a ratio of actual ice melting power to maximum ice melting power;

[0120] When the ice melting efficiency is lower than the preset threshold, the optimal ice melting power level is adjusted.

[0121] S402: Also includes remote control: establishing a communication link between a local intelligent control box and a remote short-circuit communication control box; realizing remote monitoring, fault diagnosis and emergency response.

[0122] S403: During the ice melting process, when the ambient temperature is higher than a preset temperature threshold or the ice thickness is less than a preset thickness threshold, the ice melting process is automatically stopped, and the ice melting time and power data are recorded.

[0123] In the embodiment of the present application, current monitoring is implemented using a high-precision smart meter with the following technical features: sampling frequency: 4800Hz; measurement accuracy: level 0.2; range: 0~400A; communication mode: Modbus-RTU; data refresh rate: 100ms.

[0124] Specifically, the current value judgment logic includes three thresholds: minimum threshold : A value lower than this indicates insufficient current; target threshold : The current value expected to be reached; the maximum threshold : Above this value, voltage reduction protection is required.

[0125] It should be noted that the voltage level adjustment adopts a progressive control strategy:

[0126] When the actual current I< When: If the current gear is not the highest gear, shift up a gear; if the current gear is the highest gear, trigger an alarm; the shift-up interval should not be less than 30 seconds.

[0127] when ≤I< Time: Fine-tune the voltage value of the current gear; the adjustment step is 1% of the rated value; the adjustment interval is 10 seconds.

[0128] when ≤I≤ When: Maintain the current gear; continuously monitor current fluctuations; record stable operating data.

[0129] When I> When: immediately reduce one gear; trigger overcurrent protection alarm; record overcurrent event information.

[0130] Furthermore, this embodiment also provides an ice-melting distribution transformer, comprising:

[0131] The transformer body is equipped with two sets of low-voltage output terminals, the first set of which is a normal voltage output terminal designed for power distribution users, and the second set of which is a specific voltage output terminal designed for ice melting operations;

[0132] Ice-melting control box, including local intelligent control box and remote short-circuit communication control box. The local intelligent control box is equipped with a one-touch ice-melting button.

[0133] An ice-melting power regulating device is provided at an ice-melting output end to set a plurality of gears of ice-melting power output;

[0134] The switch control system includes a boundary switch, an ice-melting switch, a protection switch and a short-circuit switch.

[0135] The second group of specific voltage output terminals has a plurality of optimal ice-melting power gears for adjusting the output power according to the actual ice covering conditions.

[0136] The local intelligent control box has a built-in deep reinforcement learning model computing unit, which is used to calculate the optimal ice-melting power level and time.

[0137] The remote short-circuit communication control box is connected to the local intelligent control box through a high-speed communication link to achieve remote synchronous control.

[0138] In summary, the present invention achieves a perfect combination of power supply and ice-melting functions by setting a structural design of dual output ends of the distribution transformer, solves the problem that traditional solutions require special ice-melting transformers or temporary modifications, and achieves the effect of one machine with two uses and saving investment.

[0139] By introducing an ice-melting control strategy based on deep reinforcement learning, the system can autonomously learn the optimal ice-melting power level and time, overcoming the defect that traditional fixed strategies cannot adapt to complex climatic conditions, realizing the intelligent and precise control of the ice-melting process, and significantly improving the ice-melting efficiency.

[0140] By designing a multi-level optimal ice-melting power gear adjustment mechanism and combining it with real-time calculation of the temperature rise sensitivity coefficient, the system can flexibly adjust the power output according to the actual icing conditions, avoiding the energy waste or insufficient ice melting caused by rough power regulation in traditional solutions, and achieving the effect of refined control, energy saving and environmental protection.

[0141] By incorporating the coordinated control strategy of boundary switches, ice-melting switches, protection switches and short-circuit switches, the system realizes fully automated operation of the ice-melting process, eliminating the safety hazards brought about by traditional manual operation, while significantly improving operating efficiency and realizing safe, reliable, efficient and convenient ice-melting operations.

[0142] By establishing a two-way real-time communication mechanism between the local intelligent control box and the remote communication control box, the system has the ability of remote monitoring, fault diagnosis and emergency handling, overcoming the problem of insufficient remote management capabilities in traditional solutions and realizing intelligent operation and maintenance and rapid response.

[0143] By designing a real-time monitoring and automatic adjustment mechanism for ice melting efficiency, the system can dynamically optimize ice melting parameters, ensure the quality of ice melting while avoiding energy waste, and achieve the unity of economy and efficiency in the ice melting process.

[0144] The present invention not only solves the problems of low efficiency, poor safety, and insufficient intelligence in traditional ice-melting solutions, but also significantly shortens the ice-melting time, greatly improves operational safety, and significantly enhances system reliability and adaptability.

[0145] Example 2 Reference Figure 3 to Figure 5 , which is the second embodiment of the present invention, and this embodiment provides an AC short-circuit ice-melting control method and an ice-melting distribution transformer.

[0146] It includes a multifunctional distribution transformer with dual functions of voltage reduction and ice melting, an ice melting control box and operation interface, and refined ice melting power regulation and ice melting control method; the multifunctional distribution transformer is equipped with two groups of low-voltage output terminals, the first group is designed for distribution users, and the second group is designed for ice melting operations; the ice melting control box consists of two parts: a local intelligent control box and a remote short-circuit communication control box; refined ice melting power regulation, and multiple gears of ice melting power output are set at the ice melting output terminal of the multifunctional distribution transformer; the ice melting control method integrates the precise control sequence of the boundary switch, ice melting switch, protection switch and short-circuit switch.

[0147] As a preferred embodiment, it includes a multifunctional distribution transformer with dual functions of voltage reduction and ice melting, an ice melting control box and an operation interface, and refined ice melting power regulation and ice melting control methods.

[0148] Preferably, the multifunctional distribution transformer is equipped with two sets of low-voltage output terminals. The first set is designed for distribution users and provides standard normal voltage output to ensure that daily electricity needs are met. The second set is designed for ice melting operations and outputs specific voltage to meet the special requirements for current and power during the ice melting process.

[0149] Preferably, the ice-melting control box consists of two parts: a local intelligent control box and a remote short-circuit communication control box. The local intelligent control box has built-in advanced control logic and is equipped with an intuitive and easy-to-use operation panel, on which a one-touch ice-melting button is specially provided; the remote short-circuit communication control box achieves remote synchronization and control with the local control box through a high-speed and stable communication link, which facilitates rapid response in emergency situations, or remote monitoring and maintenance.

[0150] As a preferred embodiment, the theoretical algorithm steps are as follows:

[0151] Step 1: Set the status assessment cycle T; obtain the equipment parameters of the ice-melting transformer and ice-melting line, including: the rated capacity S of the ice-melting transformer in operation; the power factor of the ice-melting transformer ;Number of temperature sensors in ice melting circuit ; Melting temperature boundary of ice melting line ;

[0152] Step 2: Get the real time t;

[0153] Step 3: Obtain the real-time operating data of each ice-melting transformer, which includes: the operating active power of the ice-melting transformer at the tth moment ; The measured temperature of the jth temperature sensor in the de-icing line at the tth moment ,in ;

[0154] Step 4: Obtain the historical operation data of the ice-melting transformer, which includes the active power of the ice-melting transformer at the tth moment of each of the last seven operation days. , , ; The measured temperature of the jth temperature sensor on the de-icing line at the tth time of each of the last seven operating days , , .

[0155] Step 5: Calculate the temperature rise sensitivity coefficient of each temperature sensor location on the de-icing line under different load scenarios; the temperature rise sensitivity coefficient is used to reflect the relationship between the de-icing transformer load and the temperature rise of the de-icing line. This coefficient is used to calculate the de-icing current demand in step 6, that is, by calculating the temperature that needs to be increased to achieve de-icing on the ice-covered line, and combining this coefficient to further obtain the load power that the de-icing transformer needs to achieve, and finally obtain the de-icing current demand value.

[0156] Step 5 calculates the temperature rise sensitivity coefficient of each ice-melting transformer group under different load scenarios, which refers to the following steps:

[0157] (5-1) Based on the operating active power of the de-icing transformer at the tth moment of each day in the past seven operating days , , Calculate the load rate at time t of each of the last seven operating days as follows:

[0158] ,

[0159] ,

[0160] ,

[0161] in, is the load rate at time t on the seventh operating day; is the load rate at time t on the first operating day; is the operating active power at time t on the seventh operating day; is the operating active power at time t on the first operating day; is the rated capacity of the ice melting transformer; is the power factor of the ice melting transformer.

[0162] (5-2) Divide the load rate scenarios into six gradients: [0, 20%], (20%, 40%], (40%, 60%], (60%, 80%], (80%, 100%], 100%, and label each load rate that meets the corresponding scenario. For example, if the load rate of a certain ice-melting transformer is 37.2% at a certain moment, then the ice-melting transformer is in the (20%, 40%] scenario at that moment.

[0163] Classify the distribution set F of all time points in the six load scenarios in the past seven operating days 0-20 、F 20-40 、F 40-60 、F 60-80 、F 80-100 、F> 100 , and count the number of distribution sets at each time point K 0-20 , K 20-40 , K 40-60 , K 60-80 , K 80-100 , K> 100 ;

[0164] (5-3) Calculate the active power fluctuation value of the de-icing transformer at the tth moment of each day in the last seven operating days.

[0165] Temperature fluctuation values ​​measured by each temperature sensor in the ice melting line for:

[0166] ,

[0167] ,

[0168] ,

[0169] ,

[0170] ,

[0171] ,

[0172] in, is the operating active power fluctuation value at time t on the first operating day; is the operating active power at time t on the first operating day; is the operating active power at time t-1 on the first operating day;

[0173] is the operating active power fluctuation value at time t on the 7th operating day; is the operating active power at time t on the seventh operating day; is the operating active power at time t-1 on the 7th operating day;

[0174] The temperature fluctuation value measured by the temperature sensor at time t on the first operation day of the de-icing line; The temperature measured by the temperature sensor at time t on the first operating day; The temperature measured by the temperature sensor at time t-1 on the first operating day;

[0175] The temperature fluctuation value measured by the temperature sensor at time t on the 7th operation day of the de-icing line; The temperature measured by the temperature sensor at time t on the seventh operating day; It is the temperature measured by the temperature sensor at time t-1 on the 7th operating day.

[0176] (5-4) Calculate the temperature rise sensitivity coefficient of the jth temperature sensor of the de-icing line under different load scenarios:

[0177] ,

[0178] in, is the temperature rise sensitivity coefficient of the jth temperature sensor of the de-icing line under the [0,20%]th load scenario; is the number of [0,20%] time points within 7 operating days; is the time point when the load rate is 0-20%; k is the line number; is the temperature fluctuation; is the active power fluctuation.

[0179] is the temperature rise sensitivity coefficient of the jth temperature sensor of the de-icing line under the load scenario of >100%; The number of time points >100% within 7 operating days; is the time point when the load rate is greater than 100%; k is the line number; is the temperature fluctuation; is the active power fluctuation.

[0180] Step 6: Determine the operating temperature state of the ice-melting line at the next moment and establish an ice-melting current demand calculation model based on the ice-melting line load-temperature rise test;

[0181] In step 6, judging the operating temperature state of the de-icing line at the next moment and establishing a de-icing current demand calculation model based on the de-icing line load-temperature rise test refers to the following steps:

[0182] (6-1) Based on the temperature of the ice melting line at the previous moment and the current moment, the temperature of the ice melting line at the next moment is predicted as:

[0183] ,

[0184] in, is the temperature of the jth monitoring point at the next moment; is the temperature at the jth monitoring point; is the temperature of the jth monitoring point at the previous moment.

[0185] (6-2) Calculate ice melting and cooling requirements within the range :

[0186] ,

[0187] like The line needs to be de-iced, and the line temperature needs to be increased by regulating the voltage through the de-icing transformer. ,like No ice melting is required.

[0188] (6-3) Calculate the ice melting current requirement based on the current transformer load rate and temperature rise factor:

[0189] ,

[0190] ,

[0191] In the formula, is the current value required to melt ice at line position j; The ice-melting current value that the ice-melting transformer needs to output; Accurate voltage value for ice melting transformer; According to the current load rate of the ice melting transformer Select the temperature rise sensitivity coefficient corresponding to the load rate scenario;

[0192] (6-4) Adjust the voltage level according to the current output current value of the ice-melting transformer. If the current output current value of the ice-melting transformer is , the voltage gear is adjusted up, and the boundary switch, ice-melting switch, protection switch and short-circuit switch are controlled at the same time. After adjustment, it is judged again whether the current value meets the ice-melting current requirement until the voltage gear is adjusted to make the current value meet the ice-melting current requirement.

[0193] Step 7: Wait for ΔT and return to step 2.

[0194] As a preferred option, the ice-melting power regulation is refined, and multiple levels of ice-melting power output are set at the ice-melting output end of the multi-functional distribution transformer, allowing the operator to flexibly select the appropriate power level according to the actual icing conditions, line characteristics and environmental factors.

[0195] Preferably, the ice-melting control method integrates the precise control sequence of boundary switches, ice-melting switches, protective switches and short-circuit switches; the boundary switch automatically isolates the non-ice-melting area before the ice-melting operation begins to protect other parts of the power grid from being affected; the ice-melting switch accurately controls the start-up and adjustment of the ice-melting current according to a preset program to ensure a smooth ice-melting process; the protective switch monitors the line status in real time and immediately cuts off the power supply once an abnormality is found (such as overload, short circuit, etc.) to protect the safety of equipment and personnel; the short-circuit switch controls the short-circuit of the terminal line to form an ice-melting loop.

[0196] As a preferred design formula for voltage conversion:

[0197] ,

[0198] is the voltage at the output end (normal voltage for power distribution users or voltage for ice melting); is the number of coil turns at the output end; is the number of coil turns at the input end; is the voltage at the input terminal.

[0199] As a preferred method, the ice-melting power regulation is refined, and multiple levels of ice-melting power output are set at the ice-melting output end of the multi-functional distribution transformer. The ice-melting power output is set with the following calculation formula:

[0200] ,

[0201] in, is the ice melting power; It is the voltage at the output terminal of the voltage used for ice melting; It is the current when the ice melts.

[0202] As a preferred method, the ice melting process sets the ice melting efficiency formula:

[0203] ,

[0204] in, is the ice melting efficiency; is the actual ice melting power used; It is the ice-melting power when the transformer is in the highest gear.

[0205] As a preferred method, the time parameters are controlled during the ice melting process, and the action time of the boundary switch, ice melting switch, protection switch and short-circuit switch is controlled:

[0206] ,

[0207] in, is the total ice melting time; is the action time of the boundary switch; is the action time of the ice melting switch; is the action time of the protection switch; is the action time of the short-circuit switch; is the duration of the ice melting process.

[0208] As a preferred method, the adaptive control of ice melting power based on deep reinforcement learning model is adopted to solve the problem that the variability and weak perception of various parameters in the ice melting strategy lead to difficult ice melting decision-making. In the deep reinforcement learning model, the input parameters are: ambient temperature, humidity, wind speed, ice thickness, line rated load, etc. The action strategy is affected by the predicted ice thickness. The action set is different ice melting powers, which can be discretized into 3 actions (fast, medium, slow). The ice melting action will change The value of .

[0209] (1) Calculate the predicted value of ice thickness:

[0210] ,

[0211] in, is the predicted new ice thickness (mm); is the current ice thickness (mm); is the ice thickness coefficient (usually 0.01 to 0.05 mm / (℃·%·m / s·h)); is the reference temperature (usually 0°C); is the ambient temperature (℃); is the ambient relative humidity (%); is the reference humidity (usually 85%); is the wind speed (m / s); is the time (h).

[0212] (2) Calculation of line icing load:

[0213] ,

[0214] Where ρ is the density of ice (900kg / m³); D is the diameter of the transmission line (mm); is the ice thickness (mm); L is the length of the transmission line between adjacent towers (m).

[0215] (3) Calculate the risk of pole breaking:

[0216] ,

[0217] in, is the ice load; is the rated ice load.

[0218] Calculate ice melting costs:

[0219] ,

[0220] in, For ice melting costs; is the function of power and cost; is the ice thickness (mm); the cost functions of the three actions (fast, medium, and slow) can be set based on experience.

[0221] Design reward function:

[0222] ,

[0223] in, Used to balance the impact of risk and cost; and is the weight coefficient; To reduce the risk of pole breaking; For ice melting costs.

[0224] According to the above calculations, the DQN network can be used to complete network training, and the optimal balance between ice melting power, risk and cost can be found based on actual operating data.

[0225] Example 3 Reference Figure 3 to Figure 5 , which is the third embodiment of the present invention, and this embodiment provides an AC short-circuit ice-melting control method and an ice-melting distribution transformer. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.

[0226] The specific implementation of the present invention is now further described with reference to the accompanying drawings and examples.

[0227] Obtain the equipment parameters of the ice-melting transformer and the ice-melting line. When the rated capacity S of the ice-melting transformer is 100kVA; the power factor 0.95; the number of temperature sensors in the ice melting line 1; ice melting line ice melting temperature boundary 110℃;

[0228] Obtain the real-time operation data of the ice-melting transformer, and the active power at the tth moment of the operation of the ice-melting transformer ; The measured temperature of each turbine temperature sensor at the time t .

[0229] Obtain the real-time operating data of each de-icing transformer, including the operating active power of the de-icing transformer at the tth moment. ; The measured temperature of the jth temperature sensor in the de-icing line at the tth moment ,in .

[0230] Obtain the historical operation data of the de-icing transformer. The historical operation data includes the active power of the de-icing transformer at the tth moment of each day in the past seven operation days. , , ; The measured temperature of the jth temperature sensor on the de-icing line at the tth time of each of the last seven operating days , , .

[0231] Table 1 Historical active power data

[0232] ,

[0233] Table 2 Historical load rate data

[0234] ,

[0235] Table 3 Historical temperature data

[0236] ,

[0237] Classify the distribution set F of all time points in the six load scenarios in the past seven operating days 0-20 、F 20-40 、F 40-60 、F 60-80 、F 80-100 、F> 100 , and count the number of distribution sets at each time point K 0-20 , K 20-40 , K 40-60 , K 60-80 , K 80-100 , K> 100 They are 216, 195, 152, 82, 27, and 0 respectively;

[0238] Calculate the active power fluctuation value of the de-icing transformer at the tth moment of each day in the past 7 operating days And the temperature fluctuation value measured by the temperature sensor , some data are shown in Table 4 and Table 5 below;

[0239] Table 4 Active power fluctuation data

[0240] ,

[0241] Table 5 Measured temperature fluctuation data

[0242] ,

[0243] Calculate the temperature rise sensitivity coefficient of the jth temperature sensor under different load scenarios , , , , They are 1.745, 1.417, 1.243, 0.869, and 0.694 respectively;

[0244] The real-time measurement shows that the active power of the de-icing transformer at a certain point in time is 86.7kW and the de-icing line temperature is 6.7℃. The current ice thickness is 12mm, the ice thickness coefficient is 0.01, the reference temperature is 0, the ambient temperature is 6.7℃, the ambient relative humidity is 82%, the reference humidity is 85%, the wind speed is 2m / s, the ice density is 900kg / m³, the transmission line diameter is 95mm, the transmission line length between adjacent towers is 2km, and the risk and cost weight coefficients are 0 respectively. 5 and 0.5, the active power of the previous moment was 80.3kW, and the temperature of the de-icing line at the next moment was predicted to be -3.1℃. There is a risk of icing and a need for de-icing. The load rate gradient will be between 80% and 100%, and the temperature rise coefficient is selected as 0.694. The calculated de-icing current demand value is 218.7A, and the voltage gear is adjusted up. At the same time, the boundary switch, de-icing switch, protection switch and short-circuit switch are controlled. After on-site adjustment, the de-icing current value reaches 218.7A.

[0245] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. An AC short circuit ice melting control method, characterized in that: Including, obtaining environmental parameters and equipment parameters of the ice-melting line, wherein the environmental parameters include temperature, humidity, wind speed, and ice thickness, and the equipment parameters include rated capacity, power factor, and number of temperature sensors of the ice-melting transformer; Based on the environmental parameters and equipment parameters, a deep reinforcement learning model is used to calculate the optimal ice-melting power level and ice-melting time; the deep reinforcement learning model includes: An Actor network, used for selecting an ice-melting power gear action according to a state, the Actor network comprising an input layer with the same dimension as the state, two hidden layers of 256 nodes, and an output layer corresponding to three ice-melting power gears; A critic network, used to evaluate the rationality of the selected power level, the critic network comprising an input layer of state and action information, two hidden layers of 256 nodes, and an output layer representing the Q value; According to the optimal ice-melting power level and ice-melting time, the output voltage and switch sequence of the ice-melting transformer are controlled, wherein the switch sequence includes the action timing of the boundary switch, the ice-melting switch, the protection switch and the short-circuit switch; Calculate the temperature rise sensitivity coefficient, which is calculated based on the load rate scenario division and temperature fluctuation of the ice melting transformer for 7 operating days; The operation data of the ice-melting transformer for 7 operation days are divided into six load rate scenarios according to [0, 20%], (20%, 40%], (40%, 60%], (60%, 80%], (80%, 100%], and 100%, and the temperature rise sensitivity coefficient under each scenario is calculated based on the relationship between the power fluctuation value and the temperature fluctuation value; Based on the temperature rise sensitivity coefficient, the current value in the ice melting process is monitored in real time. When the current value reaches the preset ice melting current demand value, the current voltage level is maintained; when the preset ice melting current demand value is not reached, the voltage level is adjusted until the requirement is met.

2. The AC short circuit ice melting control method according to claim 1, characterized in that: The deep reinforcement learning model is used to perform adaptive control of ice melting power: the ice melting power is discretized into three actions: fast, medium and slow; the change of ice thickness is predicted according to environmental parameters; the line ice load and the risk of pole collapse are calculated; the ice melting power is controlled based on the balance of risk and cost, and the risk and cost balance uses a reward function to control the ice melting power, wherein the reward function of the risk and cost balance needs to use the coefficient value of the ice melting cost; The prediction of ice thickness change adopts the following formula: Among them, D new is the predicted new ice thickness; d is the current ice thickness; k1 is the ice thickness coefficient; T ref is the reference temperature; T is the ambient temperature; RH is the ambient relative humidity; RH ref is the reference humidity; v is the wind speed; t is the time; The line icing load is calculated using the following formula: F=9.82×10 -8 ×ρ×π×[(D+2D new ) 2 -D 2 ]×L Where ρ is the density of ice; D is the diameter of the transmission line; D new is the ice thickness; L is the length of the transmission line between adjacent towers; The pole breaking risk P is calculated using the following formula: Where, F is the ice load; F m is the rated ice load; The ice melting cost C is calculated using the following formula: C=f (pow) ×D new Where C is the ice melting cost; f (pow) is the relationship function between power and cost; D new is the ice thickness; The reward function R that balances risk and cost reward Use the following formula: R reward =-λ1×P-λ2×C Among them, R reward Used to balance the impact of risk and cost; λ1 and λ2 are weight coefficients; P is the risk of pole breaking; C is the ice melting cost.

3. The AC short circuit ice melting control method according to claim 2, characterized in that: The control of the switching sequence includes: Before the ice melting operation begins, isolate the non-ice melting area through the boundary switch; According to the preset program, the start and adjustment of the ice melting current is controlled by the ice melting switch; Monitor the line status in real time through the protection switch and cut off the power supply when an abnormality is detected; The terminal line is short-circuited by the short-circuit switch to form an ice-melting circuit.

4. The AC short circuit ice melting control method according to claim 3, characterized in that: Also includes: Calculating and monitoring ice melting efficiency, where the ice melting efficiency is the ratio of actual ice melting power to maximum ice melting power; When the ice melting efficiency is lower than the preset threshold, the optimal ice melting power level is adjusted.

5. The AC short circuit ice melting control method according to claim 4, characterized in that: Also includes remote control: Establish a communication link through the local intelligent control box and the remote short-circuit communication control box; Realize remote monitoring, fault diagnosis and emergency response.

6. The AC short circuit ice melting control method according to claim 5, characterized in that: During the ice melting process, when the ambient temperature is higher than the preset temperature threshold or the ice thickness is less than the preset thickness threshold, the ice melting process is automatically stopped and the ice melting time and power data are recorded.

7. An ice-melting distribution transformer, based on an AC short-circuit ice-melting control method according to any one of claims 1 to 6, characterized in that: The transformer body is provided with two groups of low-voltage output terminals, the first group of which is a normal voltage output terminal designed for power distribution users, and the second group of which is a specific voltage output terminal designed for ice melting operations; The two groups of low-voltage output terminals have two states, one is to work alone, and the other is to melt ice and supply power to low-voltage users at the same time; the output voltages of the two groups of output terminals meet the following requirements: Among them, V out is the voltage at the output; N out is the number of coil turns at the output end; N in is the number of coil turns at the input end; V in is the voltage at the input terminal; The ice melting power at the specific voltage output end of the ice melting operation design satisfies: P 融冰 =V 融冰 ×I 融冰 Among them, P 融冰 is the ice melting power; V 融冰 is the voltage at the output terminal of the ice melting voltage; I 融冰 It is the current when the ice melts; An ice-melting control box, including a local intelligent control box and a remote short-circuit communication control box, wherein the local intelligent control box is provided with a one-touch ice-melting button; The ice-melting power regulating device is provided with a plurality of gears of ice-melting power output at the ice-melting output end; the efficiency of the ice-melting power output satisfies: Among them, η 融冰 is the ice melting efficiency; P 实际融冰 is the actual ice melting power used; P 最大融冰 is the ice melting power when the transformer is in the highest gear; The switch control system includes a boundary switch, an ice-melting switch, a protection switch and a short-circuit switch.

8. The ice-melting distribution transformer according to claim 7, characterized in that: The second group of specific voltage output terminals has a plurality of optimal ice-melting power gears, which are used to adjust the output voltage according to the actual ice-covered distance and the conductor model.

9. The ice-melting distribution transformer according to claim 8, characterized in that: The local intelligent control box is equipped with a deep reinforcement learning model computing unit for calculating the optimal ice-melting power level and ice-melting time.

10. The ice-melting distribution transformer according to claim 9, characterized in that: The remote short-circuit communication control box is connected to the local intelligent control box via a high-speed communication link to achieve remote synchronous control.

Images