Method, device and equipment for determining ice-melting strategy of split conductor and storage medium

Abstract

The embodiment of the application provides a kind of split conductor ice-melting strategy determination method, device, equipment and storage medium.The method first obtains the current state data of the split conductor to be ice-melted, then based on current state data, determine the thermal effect equivalent current value of the split conductor to be ice-melted, then according to the thermal effect equivalent current value and the load current value in current state data, determine the on-off duty cycle of at least one sub-conductor in the split conductor to be ice-melted, finally based on the on-off duty cycle of at least one sub-conductor and the preset ice-melting control period, determine the split conductor ice-melting strategy, wherein, the split conductor ice-melting strategy includes: the on-off time stamp of at least one sub-conductor in the split conductor to be ice-melted, the technical solution of the application can effectively improve the current regulation accuracy, and then improve the efficiency and accuracy of ice-melting.

Description

Technical Field

[0001] This application relates to the field of power transmission line technology, and in particular to a method, apparatus, equipment and storage medium for determining a de-icing strategy for split conductors. Background Technology

[0002] High-voltage transmission lines are prone to icing in extreme low temperatures, high humidity, or strong winds, leading to a sharp increase in conductor weight, mechanical stress imbalance, and even major safety accidents such as tower collapse and line breakage. Icing also significantly reduces the surface conductivity of conductors, exacerbating local corona discharge and further threatening the stable operation of the power grid. Therefore, de-icing control measures are necessary for high-voltage transmission lines.

[0003] Existing de-icing methods mainly rely on DC de-icing vehicles or manual switching. DC de-icing vehicles generate Joule heat by injecting a large current into the split conductors, forcing the ice to melt. Manual switching disconnects some of the sub-conductors in the split conductors, forcing the remaining conductors to carry a larger current to increase the heat generated for de-icing.

[0004] However, existing ice-melting methods suffer from poor current regulation precision, which affects the efficiency and accuracy of ice-melting. Summary of the Invention

[0005] This application provides a method, apparatus, device, and storage medium for determining a de-icing strategy for split conductors, in order to improve the efficiency and accuracy of de-icing.

[0006] In a first aspect, embodiments of this application provide a method for determining a de-icing strategy for split conductors, comprising:

[0007] Obtain the current status data of the split conductor to be melted;

[0008] Based on the current state data, determine the equivalent current value of the thermal effect of the ice-breaking conductor to be melted;

[0009] Based on the thermal effect equivalent current value and the load current value in the current state data, determine the duty cycle of at least one sub-conductor in the ice-melting split conductor;

[0010] Based on the duty cycle of at least one sub-conductor and the preset de-icing control cycle, a de-icing strategy for the split conductor is determined. The de-icing strategy for the split conductor includes the on / off timestamp of at least one sub-conductor in the split conductor to be de-iced.

[0011] In one or more embodiments, determining the equivalent thermal current value of the ice-breaking conductor to be melted based on the current state data includes:

[0012] Feature extraction is performed on the current state data to obtain a multidimensional feature vector;

[0013] The multidimensional feature vector is input into the random forest model to obtain the predicted melting time corresponding to at least one preset equivalent current value. The random forest model is determined based on multiple labeled state data and the labeled melting time corresponding to multiple preset equivalent current values.

[0014] The thermal effect equivalent current value of the split conductor to be melted is determined based on the shortest predicted melting time among the predicted melting times corresponding to the at least one preset equivalent current value.

[0015] In one or more embodiments, before inputting the multidimensional feature vector into a random forest model to obtain the predicted melting time corresponding to at least one preset equivalent current value, the method further includes:

[0016] Acquire the multiple labeled status data and the labeled melting time corresponding to the multiple preset equivalent current values. The labeled status data includes the historical status data of the split conductor to be melted and the simulation status data based on finite element analysis.

[0017] The initial random forest model is trained based on the labeled state data and the labeled melting time corresponding to the multiple preset equivalent current values ​​to obtain the random forest model.

[0018] In one or more embodiments, training an initial random forest model based on the labeled state data and the labeled ice-melting time corresponding to the multiple preset equivalent current values ​​to obtain the random forest model includes:

[0019] S1, For each labeled state data, feature extraction is performed on the labeled state data to obtain a labeled multidimensional feature vector;

[0020] S2, for each preset equivalent current value, the labeled multidimensional feature vector is input into the decision forest layer in the initial random forest model to obtain at least one first melting time corresponding to the preset equivalent current value. The initial random forest model includes: the decision forest layer and the fusion output layer. The decision forest layer includes at least one regression decision tree.

[0021] S3, input the at least one first ice-melting time to the fusion output layer to obtain the second ice-melting time corresponding to the preset equivalent current value;

[0022] S4. Update the parameters of the decision forest layer according to the second melting time corresponding to all preset equivalent current values ​​and all labeled melting times. Repeat steps S1-S4 until the mean square error between the second melting time corresponding to each preset equivalent current value and the corresponding labeled melting time is less than a preset threshold, and obtain the random forest model.

[0023] In one or more embodiments, determining the duty cycle of at least one sub-conductor in the de-icing splitting conductor based on the thermal effect equivalent current value and the load current value in the current state data includes:

[0024] The first ratio is determined based on the ratio between the thermal effect equivalent current value and the load current value in the current state data;

[0025] The duty cycle of the at least one sub-conductor is determined based on the square of the first ratio.

[0026] In one or more embodiments, determining the de-icing strategy for the split conductor based on the duty cycle of the at least one sub-conductor and a preset de-icing control cycle includes:

[0027] Multiply the duty cycle of the at least one sub-conductor by the preset de-icing control cycle to obtain the conduction time of the at least one sub-conductor;

[0028] Within the preset de-icing control cycle, the switching frequency is determined based on the conduction duration of the at least one sub-conductor and the preset thermal time parameter;

[0029] The at least one sub-conductor is switched on and off according to the switching frequency, and the on / off timestamp of the at least one sub-conductor is determined.

[0030] The de-icing strategy for the split conductor is determined based on the on / off timestamp of the at least one sub-conductor.

[0031] In one or more embodiments, after determining the de-icing strategy for the split conductors based on the duty cycle of the at least one sub-conductor and a preset de-icing control cycle, the method further includes:

[0032] Based on the current state data and the random forest model, determine the predicted temperature data for at least one sub-conductor;

[0033] The on / off control of at least one sub-conductor is performed according to the split conductor de-icing strategy, and the actual temperature data of at least one sub-conductor is determined.

[0034] If the difference between the predicted temperature data and the actual temperature data is greater than a preset threshold, the duty cycle of the at least one sub-conductor is updated.

[0035] Secondly, embodiments of this application provide an apparatus for determining a de-icing strategy for a split conductor, comprising:

[0036] The acquisition module is used to acquire the current status data of the split conductor to be melted.

[0037] The first determining module is used to determine the thermal effect equivalent current value of the ice-breaking conductor to be melted based on the current state data.

[0038] The second determining module is used to determine the duty cycle of at least one sub-conductor in the ice-melting splitting conductor based on the thermal effect equivalent current value and the load current value in the current state data.

[0039] The third determining module is used to determine the de-icing strategy of the split conductor based on the duty cycle of the at least one sub-conductor and the preset de-icing control cycle. The de-icing strategy of the split conductor includes the on / off timestamp of at least one sub-conductor in the split conductor to be de-iced.

[0040] In one or more embodiments, the first determining module is specifically used for:

[0041] Feature extraction is performed on the current state data to obtain a multidimensional feature vector;

[0042] The multidimensional feature vector is input into the random forest model to obtain the predicted melting time corresponding to at least one preset equivalent current value. The random forest model is determined based on multiple labeled state data and the labeled melting time corresponding to multiple preset equivalent current values.

[0043] The thermal effect equivalent current value of the split conductor to be melted is determined based on the shortest predicted melting time among the predicted melting times corresponding to the at least one preset equivalent current value.

[0044] In one or more embodiments, before inputting the multidimensional feature vector into the random forest model to obtain the predicted melting time corresponding to at least one preset equivalent current value, the first determining module is further configured to:

[0045] Acquire the multiple labeled status data and the labeled melting time corresponding to the multiple preset equivalent current values. The labeled status data includes the historical status data of the split conductor to be melted and the simulation status data based on finite element analysis.

[0046] The initial random forest model is trained based on the labeled state data and the labeled melting time corresponding to the multiple preset equivalent current values ​​to obtain the random forest model.

[0047] In one or more embodiments, the first determining module trains an initial random forest model based on the plurality of labeled state data and the labeled ice-melting time corresponding to the plurality of preset equivalent current values ​​to obtain the random forest model, specifically for:

[0048] S1, For each labeled state data, feature extraction is performed on the labeled state data to obtain a labeled multidimensional feature vector;

[0049] S2, for each preset equivalent current value, the labeled multidimensional feature vector is input into the decision forest layer in the initial random forest model to obtain at least one first melting time corresponding to the preset equivalent current value. The initial random forest model includes: the decision forest layer and the fusion output layer. The decision forest layer includes at least one regression decision tree.

[0050] S3, input the at least one first ice-melting time to the fusion output layer to obtain the second ice-melting time corresponding to the preset equivalent current value;

[0051] S4. Update the parameters of the decision forest layer according to the second melting time corresponding to all preset equivalent current values ​​and all labeled melting times. Repeat steps S1-S4 until the mean square error between the second melting time corresponding to each preset equivalent current value and the corresponding labeled melting time is less than a preset threshold, and obtain the random forest model.

[0052] In one or more embodiments, the second determining module is specifically used for:

[0053] The first ratio is determined based on the ratio between the thermal effect equivalent current value and the load current value in the current state data;

[0054] The duty cycle of the at least one sub-conductor is determined based on the square of the first ratio.

[0055] In one or more embodiments, the third determining module is specifically used for:

[0056] Multiply the duty cycle of the at least one sub-conductor by the preset de-icing control cycle to obtain the conduction time of the at least one sub-conductor;

[0057] Within the preset de-icing control cycle, the switching frequency is determined based on the conduction duration of the at least one sub-conductor and the preset thermal time parameter;

[0058] The at least one sub-conductor is switched on and off according to the switching frequency, and the on / off timestamp of the at least one sub-conductor is determined.

[0059] The de-icing strategy for the split conductor is determined based on the on / off timestamp of the at least one sub-conductor.

[0060] In one or more embodiments, after determining the de-icing strategy for the split conductor based on the duty cycle of the at least one sub-conductor and a preset de-icing control cycle, the third determining module is further configured to:

[0061] Based on the current state data and the random forest model, determine the predicted temperature data for at least one sub-conductor;

[0062] The on / off control of at least one sub-conductor is performed according to the split conductor de-icing strategy, and the actual temperature data of at least one sub-conductor is determined.

[0063] If the difference between the predicted temperature data and the actual temperature data is greater than a preset threshold, the duty cycle of the at least one sub-conductor is updated.

[0064] Thirdly, embodiments of this application provide an electronic device, including: a memory and a processor;

[0065] The memory stores computer-executed instructions;

[0066] The processor executes computer execution instructions stored in the memory, such that the processor, when executed, is used to implement the method described in the first aspect and any of the embodiments above.

[0067] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the methods described in the first aspect and any of the embodiments above.

[0068] Fifthly, this application provides a computer program product, including a computer program that, when executed by a processor, is used to implement the method for determining the split wire de-icing strategy as described in the first aspect and various possible implementations of the first aspect.

[0069] This application provides a method, apparatus, device, and storage medium for determining a de-icing strategy for a split conductor. The method first acquires the current state data of the split conductor to be de-iced. Then, based on the current state data, it determines the equivalent thermal current value of the split conductor. Next, based on the equivalent thermal current value and the load current value in the current state data, it determines the on / off duty cycle of at least one sub-conductor in the split conductor. Finally, based on the on / off duty cycle of at least one sub-conductor and a preset de-icing control cycle, it determines the de-icing strategy for the split conductor. The de-icing strategy includes: the on / off timestamps of at least one sub-conductor in the split conductor to be de-iced. In the above method, by acquiring current state data and analyzing and determining the equivalent current value of the thermal effect, the actual heat demand of the split conductor to be melted is reflected. Then, the numerical relationship between the equivalent current value of the thermal effect and the load current value is analyzed to determine the duty cycle of at least one sub-conductor. This clarifies the proportion of time each sub-conductor conducts electricity, thereby achieving precise control of the conductor temperature. Finally, by combining the duty cycle and the preset melting control cycle, a split conductor melting strategy based on the timestamp dimension is generated, clarifying the on / off timestamp of each sub-conductor in the preset melting control cycle, effectively improving melting efficiency and accuracy. Attached Figure Description

[0070] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0071] Figure 1 A flowchart illustrating the method for determining the de-icing strategy of the split conductor provided in this application embodiment. Figure 1 ;

[0072] Figure 2 A flowchart illustrating the method for determining the de-icing strategy of the split conductor provided in this application embodiment. Figure 2 ;

[0073] Figure 3 A flowchart illustrating the method for determining the de-icing strategy of the split conductor provided in this application embodiment. Figure 3 ;

[0074] Figure 4 A schematic diagram of the control device for the split conductor de-icing strategy provided in an embodiment of this application;

[0075] Figure 5 A schematic diagram of the structure of the device for determining the de-icing strategy of the split conductor provided in the embodiments of this application;

[0076] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0077] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0078] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0079] Before introducing the embodiments of this application, the application background of the embodiments of this application will be explained first:

[0080] High-voltage transmission lines are prone to icing in extreme low temperatures, high humidity, or strong winds, leading to a sharp increase in conductor weight, mechanical stress imbalance, and even major safety accidents such as tower collapse and line breakage. Icing also significantly reduces the surface conductivity of conductors, exacerbating local corona discharge and further threatening the stable operation of the power grid. Therefore, de-icing control measures are necessary for high-voltage transmission lines.

[0081] Existing de-icing methods mainly rely on DC de-icing vehicles or manual switching. DC de-icing vehicles generate Joule heat by injecting a large current into the split conductors, forcing the ice to melt. Manual switching disconnects some of the sub-conductors in the split conductors, forcing the remaining conductors to carry a larger current to increase the heat generated for de-icing.

[0082] However, the DC-based ice-melting vehicle method relies on mobile equipment for power supply, which is costly and slow to respond, making it difficult to cope with sudden icing events. The manual switch-to-off method usually uses simple binary control (fully open or fully closed), which cannot be dynamically adjusted according to real-time environmental parameters or load current, resulting in low ice-melting efficiency.

[0083] In summary, existing de-icing methods suffer from poor current regulation accuracy and uneven distribution of thermal stress in the conductors, resulting in low de-icing efficiency and accuracy.

[0084] The method for determining the de-icing strategy of split conductors provided in this application aims to solve the above-mentioned technical problems of the prior art. The technical concept of this application is as follows: Existing de-icing methods mainly regulate the current by controlling the full opening and full closing of the sub-conductors in the split conductor, resulting in low de-icing efficiency. De-icing of split conductors is mainly achieved by regulating and controlling the current of each sub-conductor in the split conductor. If the accuracy of current regulation is improved, and the on / off state and on / off duration of each sub-conductor are determined and integrated, the de-icing strategy of the split conductor can be determined, effectively improving the de-icing efficiency. Therefore, the embodiments of this application consider determining the equivalent current value of the thermal effect based on the current state data of the split conductor to be de-iced. By analyzing the numerical relationship between the equivalent current value of the thermal effect and the load current value, the on / off duty cycle of at least one sub-conductor is determined, clarifying the conduction time ratio of each sub-conductor, and realizing fine control of the conductor temperature; combined with the on / off duty cycle and the preset de-icing control cycle, a de-icing strategy of split conductors based on the timestamp dimension is generated, clarifying the on / off timestamp of each sub-conductor in the preset de-icing control cycle, effectively improving the de-icing efficiency.

[0085] The execution subject of this application embodiment is an electronic device, which can be a terminal device, such as a laptop, desktop computer, or tablet computer, or a server. In practical applications, whether the electronic device is a terminal device or a server can be determined according to the actual situation, and no specific limitation is imposed on it.

[0086] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0087] Figure 1 A flowchart illustrating the method for determining the de-icing strategy of the split conductor provided in this application embodiment. Figure 1 .like Figure 1 As shown, the method for determining the de-icing strategy of the split conductor includes the following steps:

[0088] S110. Obtain the current status data of the conductor to be melted.

[0089] In this step, the current status data of the split conductor to be melted in the actual operating environment is obtained, providing a complete and reliable data basis for the subsequent formulation of the split conductor melting strategy.

[0090] For example, the current status data includes real-time load current value, environmental parameters, and current conductor icing status data, which can comprehensively reflect the icing situation, electrical operating status, and external environmental conditions of the conductor to be melted, ensuring that the subsequent determined conductor melting strategy conforms to the actual working conditions.

[0091] Real-time load current values ​​include: the amplitude, frequency, and phase of the bus current;

[0092] Environmental parameters include: ambient temperature, relative humidity, average wind speed, instantaneous wind direction, and solar radiation intensity;

[0093] Current data on conductor icing status includes: ice thickness, ice type distribution, and ice density;

[0094] In one possible implementation, current transformers are installed on the split conductors and towers to collect the real-time load current value of the split conductors to be melted, environmental parameters are collected based on micro-weather stations, and the current icing status data of the conductors is monitored in real time by combining image recognition technology or tension sensors.

[0095] S120. Based on the current state data, determine the equivalent current value of the thermal effect of the conductor to be melted.

[0096] In this step, using the current state data as input and combining it with the Joule heating characteristics required for ice melting, the equivalent current of the thermal effect that can achieve the expected ice melting effect within a safe range is determined.

[0097] For example, the thermal effect equivalent current value, with thermal effect equivalence as its core, characterizes the equivalent current amplitude required to achieve the target ice melting power.

[0098] In one possible implementation, a multi-dimensional feature vector containing multi-source information is constructed based on real-time load current values, environmental parameters, and current conductor icing status data. This multi-dimensional feature vector is then input into a pre-trained ensemble learning model (such as a random forest regression model). By performing online inference calculations on the gradients of different equivalent current values, the mapping relationship between different equivalent current values ​​and the predicted melting time is output. Finally, by combining the melting time requirements and conductor thermal limit constraints (such as the safe extreme value of conductor material annealing temperature), the thermal effect equivalent current value is determined.

[0099] S130. Based on the thermal effect equivalent current value and the load current value in the current state data, determine the duty cycle of at least one sub-conductor in the de-icing split conductor.

[0100] In this step, based on the ratio between the thermal effect equivalent current value and the load current value in the current state data, the duty cycle of at least one sub-conductor in the ice-melting split conductor is calculated to achieve a precise match between the target heating power and the actual heating power.

[0101] In one possible implementation, the ratio of the thermal effect equivalent current to the load current is first calculated. Then, according to Joule's law, the duty cycle is defined as the square of this current ratio. The total duty cycle is then evenly distributed to each sub-conductor to obtain the on / off duty cycle of each sub-conductor during the control cycle.

[0102] In one possible implementation, step S130 described above may include the following steps:

[0103] Step 1: Determine the first ratio based on the ratio between the thermal effect equivalent current value and the load current value in the current state data;

[0104] For example, the ratio between the thermal effect equivalent current value and the load current value in the current state data is calculated to obtain the first ratio, which intuitively reflects the proportional relationship between the current required for the target heating and the actual available current.

[0105] Step 2: Determine the duty cycle of at least one sub-conductor based on the square of the first ratio.

[0106] For example, based on the physical property that the heat generation is proportional to the square of the current in Joule's law, the first ratio is squared to obtain the total duty cycle required for the entire split conductor. This is then distributed to individual sub-conductors to finally determine the duty cycle of each sub-conductor, ensuring that the total heat generation within the cycle is consistent with the heat generated by the thermal effect equivalent current value.

[0107] In one possible implementation, the total on / off duty cycle The calculation formula is as follows:

[0108]

[0109] in, This represents the equivalent current value due to the thermal effect. This indicates the load current value in the current status data.

[0110] S140. Determine the de-icing strategy for the split conductor based on the duty cycle of at least one sub-conductor and the preset de-icing control cycle.

[0111] The de-icing strategy for split conductors includes: the on / off timestamp of at least one sub-conductor in the split conductor to be de-iced.

[0112] In this step, based on the duty cycle of at least one sub-conductor and the preset de-icing control cycle, a split conductor de-icing strategy that can be directly executed is generated. The on and off times of each sub-conductor are precisely controlled by the on and off timestamps. Within a complete preset de-icing control cycle, the average thermal effect and the thermal effect equivalent current value are consistent, while ensuring that the heating of each sub-conductor is balanced.

[0113] In one possible implementation, the conduction duration and off duration of a single sub-conductor are calculated based on the duty cycle of the single sub-conductor and the preset de-icing control cycle. The conduction periods of each sub-conductor are staggered within the preset de-icing control cycle to avoid current surges caused by simultaneous on / off. Spatial rotation logic is executed within a continuous cycle to generate a sequence of on / off timestamps containing the on start time and off time of each sub-conductor, thus forming the final split conductor de-icing strategy.

[0114] For example, in the case of four-split conductors, if de-icing requires 75% power output, then in four consecutive preset de-icing control cycles, sub-conductors A, B, C, and D are sequentially controlled to undertake 25% of the disconnection task. By spatially rotating, it is ensured that there is 75% effective power output in each preset de-icing control cycle. This not only ensures that the average heat input of each sub-conductor is completely equal, eliminating the temperature difference between the split conductors in space, but also avoids secondary severe icing or mechanical imbalance caused by long-term disconnection of a single conductor.

[0115] The method for determining the de-icing strategy of the split conductor provided in this application embodiment first obtains the current state data of the split conductor to be de-iced, then determines the thermal effect equivalent current value of the split conductor to be de-iced based on the current state data, then determines the on / off duty cycle of at least one sub-conductor in the split conductor to be de-iced based on the thermal effect equivalent current value and the load current value in the current state data, and finally determines the de-icing strategy of the split conductor based on the on / off duty cycle of at least one sub-conductor and the preset de-icing control cycle. The de-icing strategy of the split conductor includes: the on / off timestamp of at least one sub-conductor in the split conductor to be de-iced. In this embodiment, by acquiring current state data and analyzing and determining the equivalent current value of the thermal effect, the actual heat demand of the split conductor to be melted is reflected. Then, the numerical relationship between the equivalent current value of the thermal effect and the load current value is analyzed to determine the duty cycle of at least one sub-conductor. This clarifies the proportion of time each sub-conductor conducts electricity, thereby achieving precise control of the conductor temperature. Finally, by combining the duty cycle and the preset melting control cycle, a split conductor melting strategy based on the timestamp dimension is generated, clarifying the on / off timestamp of each sub-conductor in the preset melting control cycle, effectively improving melting efficiency and accuracy.

[0116] Based on the above embodiments, Figure 2 A flowchart illustrating the method for determining the de-icing strategy of the split conductor provided in this application embodiment. Figure 2 .like Figure 2 As shown, one possible implementation of step S120 above includes the following steps:

[0117] S210. Extract features from the current state data to obtain a multidimensional feature vector.

[0118] In this step, the current state data is structured and normalized to remove redundant information and extract key influencing factors, forming a multidimensional feature vector suitable for input to the ensemble learning model.

[0119] In one possible implementation, the raw data such as real-time load current value, ice thickness, ambient temperature, and relative humidity in the current state data are numerically normalized and dimensionally aligned. Discrete and continuous data are uniformly encoded for feature extraction and combined to form a multi-dimensional feature vector containing electrical features, environmental features, and ice features. This satisfies the input format requirements of the random forest model and improves the accuracy and stability of subsequent model inference.

[0120] S220. Input the multidimensional feature vector into the random forest model to obtain the predicted ice melting time corresponding to at least one preset equivalent current value;

[0121] The random forest model is determined based on multiple labeled state data and the labeled melting time corresponding to multiple preset equivalent current values.

[0122] In this step, a random forest model trained based on multiple labeled state data and labeled melting times corresponding to multiple preset equivalent current values ​​is used for inference and prediction. The multidimensional feature vector is input into the random forest model to quickly predict the melting time under different preset equivalent current values, and outputs at least one predicted melting time corresponding to a preset equivalent current value.

[0123] In one possible implementation, multiple preset equivalent current values ​​with gradient distributions are pre-defined as candidates. The multidimensional feature vector is input into a random forest model, which then performs inference calculations on each candidate equivalent current in sequence and outputs the corresponding predicted melting time. This yields a mapping relationship between the "preset equivalent current value and the predicted melting time". Based on this mapping relationship, the equivalent current value corresponding to the water film that can be generated at the interface between the ice layer and the conductor most quickly enough to cause the ice to fall off can be quickly determined.

[0124] In one possible implementation, prior to step S220 above, the method for determining the de-icing strategy for the split conductor further includes the following steps:

[0125] S21. Obtain multiple labeled status data and multiple preset equivalent current values ​​corresponding to the labeled ice melting time;

[0126] The labeled status data includes historical status data of the ice-breaking conductor and simulation status data based on finite element analysis.

[0127] For example, historical state data of the split conductor to be melted is collected and combined with simulation state data based on finite element analysis to construct a labeled dataset for training a random forest model. The labeled dataset includes multiple labeled state data and labeled melting times corresponding to multiple preset equivalent current values.

[0128] In one possible implementation, historical status data is collected from the transmission line monitoring platform. Actual operating data such as conductor load current and ambient temperature and humidity can be selected under different seasons and different ice thicknesses. Furthermore, a corresponding preset equivalent current and labeled ice melting time are matched for each set of historical status data.

[0129] A simulation model of heat conduction for ice melting on conductors was constructed based on finite element analysis. The simulation environment included an ambient temperature step from -25℃ to 0℃, a wind speed step from 0m / s to 25m / s, and a humidity step from 10% to 95%. By simulating the ice melting process under different temperature and humidity, wind speed, and ice thickness conditions, simulation state data was obtained, and a corresponding preset equivalent current and labeled ice melting time were matched for each set of simulation state data.

[0130] Combining historical state data with simulated state data expands the diversity and coverage of the data, providing reliable sample support for subsequent training of high-precision random forest models.

[0131] S22. Train the initial random forest model based on multiple labeled state data and the labeled melting time corresponding to multiple preset equivalent current values ​​to obtain the random forest model.

[0132] For example, the initial random forest model can be trained using a pre-constructed labeled dataset to learn the mapping rules between different features, ultimately resulting in a random forest model with accurate online prediction capabilities.

[0133] In one possible implementation, the labeled dataset containing historical state data and simulation state data is divided into a training set and a validation set in proportion. The initial random forest model is iteratively trained using the training set, and the decision tree splitting parameters of the initial random forest model are continuously adjusted until the model determines the decision tree splitting threshold based on the principle of minimizing mean squared error, at which point training stops.

[0134] In one possible implementation, step S22 described above may include the following steps:

[0135] S1, For each labeled state data, perform feature extraction on the labeled state data to obtain a labeled multidimensional feature vector;

[0136] For example, the collected labeled status data is standardized and key information is extracted, transforming the original working condition information represented by the labeled status data into a structured labeled multidimensional feature vector suitable for model input, providing a standardized and unified data foundation for subsequent model training.

[0137] In one possible implementation, the labeled state data, which includes historical state data and simulation state data based on finite element analysis, is numerically normalized according to multiple dimensions such as load current, ice thickness, ambient temperature, and wind speed to remove dimensional differences and form a labeled multidimensional feature vector in a unified format.

[0138] S2, for each preset equivalent current value, the labeled multidimensional feature vector is input into the decision forest layer in the initial random forest model to obtain at least one first melting time corresponding to the preset equivalent current value;

[0139] The initial random forest model includes a decision forest layer and a fusion output layer, wherein the decision forest layer includes at least one regression decision tree;

[0140] For example, the labeled multidimensional feature vector is input into the decision forest layer in the initial random forest model. Multiple regression decision trees in the decision forest layer are used to perform parallel reasoning on the labeled multidimensional feature vector. Each regression decision tree performs recursive judgment from the root node to the leaf node to fit the nonlinear mapping between the Joule heat generation inside the conductor and the external power dissipation. Finally, an initial melting time, i.e. the first melting time, is output independently.

[0141] In one possible implementation, the decision forest layer of the initial random forest model makes branch judgments based on the preset equivalent current value and the labeled multidimensional feature vector. It uses a bootstrap sampling method to randomly extract independent subsets from the labeled multidimensional feature vector and assigns a subset to each regression decision tree in the decision forest layer. It also randomly selects feature subsets for node splitting. Each regression decision tree independently completes forward inference and outputs the corresponding first melting time.

[0142] S3, input at least one first melting time to the fusion output layer to obtain a second melting time corresponding to a preset equivalent current value;

[0143] For example, the fusion output layer performs a weighted average calculation on the multiple first ice-melting times output by multiple regression decision trees to eliminate the prediction bias of individual trees and obtain a stable and unified model output result, namely the second ice-melting time.

[0144] S4. Update the parameters of the decision forest layer according to the second melting time corresponding to all preset equivalent current values ​​and all labeled melting times. Repeat steps S1-S4 until the mean square error between the second melting time corresponding to each preset equivalent current value and the corresponding labeled melting time is less than the preset threshold, and obtain the random forest model.

[0145] For example, the mean square error between the second melting time output by the initial random forest model and the labeled melting time is used as the optimization objective, and the splitting threshold of the decision forest layer is iteratively updated until the mean square error is less than the preset threshold.

[0146] In one possible implementation, the mean squared error between the second melting time and the labeled melting time is calculated. If the mean squared error is greater than a preset threshold, the splitting features and thresholds of each regression decision tree are updated in reverse. After repeating multiple rounds of training, the error is reduced to within the threshold, the training ends, and a converged random forest model is obtained.

[0147] During the growth of each regression decision tree, random samples are taken. We have candidate feature vectors, and based on the principle of minimizing mean square error, we find the optimal splitting attribute and splitting threshold.

[0148] The formula for calculating the mean squared error is as follows:

[0149]

[0150] in, Indicates the first The preset equivalent current value corresponds to the marked ice-melting time. Indicates the first The second melting time corresponds to a preset equivalent current value. This represents the mean square error between the second melting time and the marked melting time.

[0151] Furthermore, in one possible implementation, the training of the initial random forest model described above can deeply learn the nonlinear coupling relationships of the parameters in the conductor heat balance equation, effectively identifying the heat dissipation area under different wind speeds. Overall quality and specific heat capacity The contribution weights to the temperature rise trajectory. This physical process follows the improved conductor heat balance equation:

[0152]

[0153] in, This represents the equivalent effective current of the conductor. This indicates the convective heat transfer coefficient affected by wind speed. and These represent the surface temperature of the conductor and the ambient temperature, respectively. Indicates emissivity. Represents the Stefan-Boltzmann constant. This indicates the latent heat of ice melting.

[0154] By learning the weights of each variable in the above physical equations, the random forest model can output the optimal ice-melting current value without the need to build a complex analytical model.

[0155] S230. Determine the thermal effect equivalent current value of the conductor to be melted based on the shortest predicted melting time among the predicted melting times corresponding to at least one preset equivalent current value.

[0156] In this step, with the goal of optimizing the de-icing efficiency, the shortest predicted de-icing time is determined from the predicted de-icing time corresponding to at least one preset equivalent current value, and the corresponding preset equivalent current is determined as the thermal effect equivalent current value, so as to achieve the fastest de-icing while ensuring the safe operation of the conductor.

[0157] In one possible implementation, all preset equivalent current values ​​and their corresponding predicted melting times are iterated through, the shortest predicted melting time is selected by comparison, and the preset equivalent current value corresponding to this time is taken as the final thermal effect equivalent current value. If there are multiple preset equivalent current values ​​corresponding to similar shortest predicted melting times, the best value is selected by combining the conductor heat capacity constraint to ensure that the melting speed and operational safety are taken into account.

[0158] The method for determining the de-icing strategy for split conductors provided in this application first extracts features from the current state data to obtain a multi-dimensional feature vector. Then, the multi-dimensional feature vector is input into a random forest model to obtain the predicted de-icing time corresponding to at least one preset equivalent current value. The random forest model is determined based on multiple labeled state data and labeled de-icing times corresponding to multiple preset equivalent current values. Next, the thermal effect equivalent current value of the split conductor to be de-iced is determined based on the shortest predicted de-icing time among the predicted de-icing times corresponding to at least one preset equivalent current value. In this embodiment, feature extraction of the current state data transforms complex data into a multi-dimensional feature vector, providing a data foundation for subsequent model processing. By introducing a random forest model, the nonlinear mapping relationship between different features in the multi-dimensional feature vector is deeply learned, accurately predicting the de-icing time corresponding to different preset equivalent current values, thus solving the problem of insufficient dynamic environment adaptability in traditional de-icing control. Finally, the equivalent current value corresponding to the shortest predicted de-icing time is selected to determine the thermal effect equivalent current value of the split conductor to be de-iced, ensuring de-icing efficiency while avoiding overheating of the conductor and energy waste.

[0159] Based on the above embodiments, Figure 3 A flowchart illustrating the method for determining the de-icing strategy of the split conductor provided in this application embodiment. Figure 3 .like Figure 3 As shown, one possible implementation of step S130 above includes the following steps:

[0160] S310. Multiply the duty cycle of at least one sub-conductor by the preset de-icing control cycle to obtain the conduction time of at least one sub-conductor.

[0161] In this step, the on / off duty cycle is quantified into a specific time length in the preset ice-melting control cycle by multiplying the on / off duty cycle with the control cycle. This clarifies the actual duration for which a single sub-conductor needs to remain conductive within a complete preset ice-melting control cycle, providing a quantitative time basis for subsequent time-domain on / off control.

[0162] In one possible implementation, for multiple sub-conductors in the de-icing split conductor, the conduction time corresponding to each conductor is calculated separately to ensure that the total heat generated by each sub-conductor during the conduction time is equal to the heat generated by the thermal effect current value within the preset de-icing control cycle.

[0163] S320. Within the preset de-icing control cycle, determine the switching frequency based on the conduction duration of at least one sub-conductor and the preset thermal time parameter.

[0164] In this step, by combining the preset thermal time parameter related to the conduction duration and the thermal inertia of the conductor, a reasonable switching frequency for at least one sub-conductor within the preset de-icing control cycle is determined. This ensures both uniform and stable heating and avoids excessive wear on the switching devices or drastic temperature fluctuations caused by frequent switching.

[0165] In one possible implementation, the length of the preset de-icing control cycle is dynamically adjusted by calculating the preset thermal time parameters of the split conductor (such as the thermal time constant based on conductor material parameters and environmental parameters), thereby determining the switching frequency.

[0166] For example, in low-temperature and high-humidity environments, the thermal time constant is small, requiring a high-frequency control cycle of 1 second to quickly respond to temperature fluctuations; while in high-temperature and low-humidity environments, the thermal time constant is large, allowing the control cycle to be extended to 10 seconds to reduce equipment wear.

[0167] S330. Perform on / off switching on at least one sub-conductor according to the switching frequency, and determine the on / off timestamp of at least one sub-conductor.

[0168] In this step, based on the switching frequency, the conduction duration of at least one sub-conductor is segmented and arranged within the preset de-icing control cycle, generating a precise timestamp including the start time of conduction and the turn-off time, forming a time-domain on / off sequence that can be directly executed, ensuring that the total conduction duration within the preset de-icing control cycle is consistent with the target value.

[0169] In one possible implementation, the disconnection tasks of different sub-conductors are rotated in each preset de-icing control cycle according to the switching frequency, ensuring that the average energization time of all sub-conductors is equal. By dynamically adjusting the disconnection tasks, secondary icing caused by long-term power outages or thermal aging caused by excessive energization of a single conductor is avoided, while maintaining the overall thermal stress balance of the conductor.

[0170] S340. Determine the de-icing strategy for split conductors based on the on / off timestamps of at least one sub-conductor.

[0171] In this step, the on / off timestamps of all sub-conductors are integrated to generate a complete and executable split conductor de-icing strategy. This split conductor de-icing strategy can precisely control the on / off sequence of each sub-conductor, achieve the target thermal effect within the preset de-icing control cycle, and ensure that the heating of each sub-conductor is balanced.

[0172] In one possible implementation, after the aforementioned step S140, the method for determining the de-icing strategy for the split conductor further includes the following steps:

[0173] Step 1: Based on the current state data and the random forest model, determine the predicted temperature data for at least one sub-wire;

[0174] For example, features are extracted from the current state data and input into a random forest model to obtain the mapping relationship between multiple preset equivalent current values ​​and the predicted ice melting time. A target preset equivalent current value is applied to at least one sub-conductor, and the curve data of the temperature change of at least one sub-conductor over time during the predicted ice melting time is simulated, i.e., the predicted temperature data.

[0175] Step 2: Control the on / off state of at least one sub-conductor according to the split conductor de-icing strategy and determine the actual temperature data of at least one sub-conductor;

[0176] For example, at least one sub-conductor is switched on and off according to the generated split conductor de-icing strategy, and the surface temperature of the conductor during operation is collected by a sensing device, and the curve data of the conductor surface temperature changing with time during the entire operation is recorded, i.e., the real temperature data.

[0177] Step 3: If the difference between the predicted temperature data and the actual temperature data is greater than the preset threshold, then the duty cycle of at least one sub-conductor is updated.

[0178] For example, the predicted temperature data is compared with the actual temperature data. If the difference between the temperature rise slope or the average temperature between the predicted temperature data and the actual temperature data is greater than a preset threshold (preset slope threshold or preset temperature difference threshold), the duty cycle of at least one sub-conductor is dynamically adjusted according to the magnitude of the difference.

[0179] Furthermore, in one possible implementation, Figure 4 A schematic diagram of the control device for the split conductor de-icing strategy provided in an embodiment of this application. Figure 4 As shown, the control device for the split conductor de-icing strategy includes: a sensor array 10 installed at the power transmission site, a central intelligent control unit 20, a solid-state circuit breaker array 30 deployed on the sub-conductors, and a split conductor 40.

[0180] Sensor array 10 is installed at key stress points or high-level crossarms of transmission towers in the power transmission site. It integrates meteorological sensors, an icing monitoring unit, and a current transformer to collect current status data of the split conductors to be melted. Among them, the meteorological sensors are used to acquire environmental parameters around the conductors in real time, including ambient temperature, relative humidity, and average wind speed; the icing monitoring unit uses a high-precision tension sensor or image recognition camera to acquire the ice thickness, ice pattern distribution, and ice density of the split conductors 40 in real time; and the current transformer is connected in series on the busbar side to acquire the real-time load current value.

[0181] The central intelligent control unit 20 is connected to the sensor array 10 via wired or wireless communication. The central intelligent control unit 20 integrates a data preprocessing unit, a random forest inference engine, and a time-domain sequence generator. The data preprocessing unit performs filtering, denoising, and normalization on the collected current state data, constructing a multi-dimensional feature vector that meets the input requirements of the random forest model. The random forest inference engine, composed of pre-trained random forest model parameters, is responsible for determining the equivalent thermal current value and on / off duty cycle based on the multi-dimensional feature vector. The time-domain sequence generator converts the calculated on / off duty cycle into a specific pulse control sequence, generating a split-conductor ice-melting strategy.

[0182] The solid-state circuit breaker array 30 consists of multiple independent solid-state circuit breaker units, each connected in series in the sub-circuit of the split conductor 40. Each solid-state circuit breaker unit employs a power electronic semiconductor switching architecture, with its core switching device being an array of insulated-gate bipolar transistors (IGBTs) or silicon carbide (SiC) modules capable of handling high voltage and high current, providing microsecond-level fast breaking capability. The solid-state circuit breaker unit is also equipped with an inductive power extraction device to sense electrical energy from the high-voltage conductors and store it in a capacitor to power the circuit breaker's drive control circuit. Furthermore, each solid-state circuit breaker unit receives control commands from the central intelligent control unit 20 via a wireless communication module, enabling coordinated switching.

[0183] The method for determining the split conductor de-icing strategy provided in this application embodiment first multiplies the duty cycle of at least one sub-conductor by a preset de-icing control period to obtain the conduction duration of at least one sub-conductor. Then, within the preset de-icing control period, a switching frequency is determined based on the conduction duration of at least one sub-conductor and a preset thermal time parameter. Next, the at least one sub-conductor is switched on and off according to the switching frequency, and the on / off timestamps of at least one sub-conductor are determined. Finally, the split conductor de-icing strategy is determined based on the on / off timestamps of at least one sub-conductor. In this embodiment, multiplying the duty cycle of the sub-conductor by the preset de-icing control period allows for accurate calculation of the conduction duration of at least one sub-conductor. Then, combined with the preset thermal time parameter, a reasonable switching frequency is determined to dynamically adjust the on / off state of at least one sub-conductor, ensuring that the heat output during the de-icing process matches the actual demand. By recording the on / off timestamps of at least one sub-conductor, a split conductor de-icing strategy that can precisely control the on / off state of the sub-conductors is generated, improving de-icing efficiency.

[0184] Based on the above embodiments, the following are embodiments of the apparatus involved in this application:

[0185] Figure 5 This is a schematic diagram of the device for determining the de-icing strategy of the split conductor provided in an embodiment of this application. Figure 5 As shown, the device 500 for determining the de-icing strategy of the split conductor includes:

[0186] The acquisition module 510 is used to acquire the current status data of the split conductor to be melted;

[0187] The first determining module 520 is used to determine the equivalent current value of the thermal effect of the ice-breaking splitting conductor based on the current state data;

[0188] The second determining module 530 is used to determine the duty cycle of at least one sub-conductor in the ice-melting split conductor based on the thermal effect equivalent current value and the load current value in the current state data.

[0189] The third determining module 540 is used to determine the de-icing strategy of the split conductor based on the duty cycle of at least one sub-conductor and the preset de-icing control cycle. The de-icing strategy of the split conductor includes the on / off timestamp of at least one sub-conductor in the split conductor to be de-iced.

[0190] In one or more embodiments, the first determining module 520 is specifically used for:

[0191] Feature extraction is performed on the current state data to obtain a multidimensional feature vector;

[0192] By inputting multidimensional feature vectors into a random forest model, the predicted melting time corresponding to at least one preset equivalent current value is obtained. The random forest model is determined based on multiple labeled state data and the labeled melting time corresponding to multiple preset equivalent current values.

[0193] The thermal effect equivalent current value of the split conductor to be melted is determined based on the shortest predicted melting time among the predicted melting times corresponding to at least one preset equivalent current value.

[0194] In one or more embodiments, before inputting the multidimensional feature vector into the random forest model to obtain the predicted melting time corresponding to at least one preset equivalent current value, the first determining module 520 is further configured to:

[0195] Acquire multiple labeled status data and labeled melting times corresponding to multiple preset equivalent current values. The labeled status data includes historical status data of the conductor to be melted and simulation status data based on finite element analysis.

[0196] The initial random forest model is trained based on multiple labeled state data and labeled melting times corresponding to multiple preset equivalent current values ​​to obtain the random forest model.

[0197] In one or more embodiments, the first determining module 520 trains an initial random forest model based on multiple labeled state data and labeled melting times corresponding to multiple preset equivalent current values ​​to obtain a random forest model, specifically used for:

[0198] S1, For each labeled state data, perform feature extraction on the labeled state data to obtain a labeled multidimensional feature vector;

[0199] S2, for each preset equivalent current value, the labeled multidimensional feature vector is input into the decision forest layer in the initial random forest model to obtain at least one first melting time corresponding to the preset equivalent current value. The initial random forest model includes: a decision forest layer and a fusion output layer. The decision forest layer includes at least one regression decision tree.

[0200] S3, input at least one first melting time to the fusion output layer to obtain a second melting time corresponding to a preset equivalent current value;

[0201] S4. Update the parameters of the decision forest layer according to the second melting time corresponding to all preset equivalent current values ​​and all labeled melting times. Repeat steps S1-S4 until the mean square error between the second melting time corresponding to each preset equivalent current value and the corresponding labeled melting time is less than the preset threshold, and obtain the random forest model.

[0202] In one or more embodiments, the second determining module 530 is specifically used for:

[0203] The first ratio is determined based on the ratio between the thermal effect equivalent current value and the load current value in the current state data;

[0204] Based on the square of the first ratio, determine the duty cycle of at least one sub-conductor.

[0205] In one or more embodiments, the third determining module 540 is specifically used for:

[0206] Multiply the duty cycle of at least one sub-conductor by the preset de-icing control cycle to obtain the conduction time of at least one sub-conductor;

[0207] Within the preset de-icing control cycle, the switching frequency is determined based on the conduction duration of at least one sub-conductor and the preset thermal time parameter;

[0208] At least one sub-conductor is switched on and off according to the switching frequency, and the on and off timestamps of at least one sub-conductor are determined.

[0209] Determine the de-icing strategy for split conductors based on the on / off timestamps of at least one sub-conductor.

[0210] In one or more embodiments, after determining the de-icing strategy for the split conductor based on the on / off duty cycle of at least one sub-conductor and a preset de-icing control cycle, the third determining module 540 is further configured to:

[0211] Based on the current state data and the random forest model, determine the predicted temperature data for at least one sub-wire;

[0212] Based on the split conductor de-icing strategy, the on / off control of at least one sub-conductor is performed and the actual temperature data of at least one sub-conductor is determined;

[0213] If the difference between the predicted temperature data and the actual temperature data is greater than a preset threshold, the duty cycle of at least one sub-conductor will be updated.

[0214] The device for determining the ice-melting strategy of the split conductor provided in this embodiment can execute the method provided in the above method embodiment. Its implementation principle and technical effect are similar, and will not be described in detail here.

[0215] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 6 As shown, the electronic device 600 includes: a processor 610, a memory 620, and a bus 630;

[0216] The memory 620 is used to store the computer-executed instructions of the processor 610;

[0217] The processor 610 is configured to execute the technical solutions of any of the foregoing method embodiments by executing computer execution instructions.

[0218] The specific implementation process of processor 610 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.

[0219] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0220] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.

[0221] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0222] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.

[0223] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the above-described method.

[0224] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random-Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read-Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0225] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an application-specific integrated circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in the device.

[0226] The division of units is merely a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0227] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0228] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0229] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory, random access memory (RAM), magnetic disks, or optical disks.

[0230] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0231] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

Claims

1. A method for determining a de-icing strategy for split conductors, characterized in that, include: Obtain the current status data of the split conductor to be melted; Based on the current state data, determine the equivalent current value of the thermal effect of the ice-breaking conductor to be melted; Based on the thermal effect equivalent current value and the load current value in the current state data, determine the duty cycle of at least one sub-conductor in the ice-melting split conductor; Based on the duty cycle of at least one sub-conductor and the preset de-icing control cycle, a de-icing strategy for the split conductor is determined. The de-icing strategy for the split conductor includes the on / off timestamp of at least one sub-conductor in the split conductor to be de-iced.

2. The method according to claim 1, characterized in that, Determining the equivalent current value of the thermal effect of the ice-breaking conductor based on the current state data includes: Feature extraction is performed on the current state data to obtain a multidimensional feature vector; The multidimensional feature vector is input into the random forest model to obtain the predicted melting time corresponding to at least one preset equivalent current value. The random forest model is determined based on multiple labeled state data and the labeled melting time corresponding to multiple preset equivalent current values. The thermal effect equivalent current value of the split conductor to be melted is determined based on the shortest predicted melting time among the predicted melting times corresponding to the at least one preset equivalent current value.

3. The method according to claim 2, characterized in that, Before inputting the multidimensional feature vector into the random forest model to obtain the predicted melting time corresponding to at least one preset equivalent current value, the method further includes: Acquire the multiple labeled status data and the labeled melting time corresponding to the multiple preset equivalent current values. The labeled status data includes the historical status data of the split conductor to be melted and the simulation status data based on finite element analysis. The initial random forest model is trained based on the labeled state data and the labeled melting time corresponding to the multiple preset equivalent current values ​​to obtain the random forest model.

4. The method according to claim 3, characterized in that, The step of training an initial random forest model based on the multiple labeled state data and the labeled ice-melting time corresponding to the multiple preset equivalent current values ​​to obtain the random forest model includes: S1, For each labeled state data, feature extraction is performed on the labeled state data to obtain a labeled multidimensional feature vector; S2, for each preset equivalent current value, the labeled multidimensional feature vector is input into the decision forest layer in the initial random forest model to obtain at least one first melting time corresponding to the preset equivalent current value. The initial random forest model includes: the decision forest layer and the fusion output layer. The decision forest layer includes at least one regression decision tree. S3, input the at least one first ice-melting time to the fusion output layer to obtain the second ice-melting time corresponding to the preset equivalent current value; S4. Update the parameters of the decision forest layer according to the second melting time corresponding to all preset equivalent current values ​​and all labeled melting times. Repeat steps S1-S4 until the mean square error between the second melting time corresponding to each preset equivalent current value and the corresponding labeled melting time is less than a preset threshold, and obtain the random forest model.

5. The method according to any one of claims 1-4, characterized in that, The step of determining the duty cycle of at least one sub-conductor in the de-icing splitting conductor based on the equivalent current value of the thermal effect and the load current value in the current state data includes: The first ratio is determined based on the ratio between the thermal effect equivalent current value and the load current value in the current state data; The duty cycle of the at least one sub-conductor is determined based on the square of the first ratio.

6. The method according to claim 5, characterized in that, The method of determining the de-icing strategy for the split conductors based on the duty cycle of at least one sub-conductor and a preset de-icing control cycle includes: Multiply the duty cycle of the at least one sub-conductor by the preset de-icing control cycle to obtain the conduction time of the at least one sub-conductor; Within the preset de-icing control cycle, the switching frequency is determined based on the conduction duration of the at least one sub-conductor and the preset thermal time parameter; The at least one sub-conductor is switched on and off according to the switching frequency, and the on / off timestamp of the at least one sub-conductor is determined. The de-icing strategy for the split conductor is determined based on the on / off timestamp of the at least one sub-conductor.

7. The method according to any one of claims 2-4, characterized in that, After determining the de-icing strategy for the split conductors based on the duty cycle of at least one sub-conductor and a preset de-icing control cycle, the method further includes: Based on the current state data and the random forest model, determine the predicted temperature data for at least one sub-conductor; The on / off control of at least one sub-conductor is performed according to the split conductor de-icing strategy, and the actual temperature data of at least one sub-conductor is determined. If the difference between the predicted temperature data and the actual temperature data is greater than a preset threshold, the duty cycle of the at least one sub-conductor is updated.

8. A device for determining a split conductor de-icing strategy, characterized in that, include: The acquisition module is used to acquire the current status data of the split conductor to be melted. The first determining module is used to determine the thermal effect equivalent current value of the ice-breaking conductor to be melted based on the current state data. The second determining module is used to determine the duty cycle of at least one sub-conductor in the ice-melting splitting conductor based on the thermal effect equivalent current value and the load current value in the current state data. The third determining module is used to determine the de-icing strategy of the split conductor based on the duty cycle of the at least one sub-conductor and the preset de-icing control cycle. The de-icing strategy of the split conductor includes the on / off timestamp of at least one sub-conductor in the split conductor to be de-iced.

9. An electronic device, characterized in that, include: Memory, processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory, causing the processor to perform the method as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1-7.

Images