Temperature estimation model construction method and device, computer equipment and storage medium

Abstract

The invention discloses a temperature estimation model construction method and device, computer equipment and a storage medium, relates to the field of motor control, and is used for solving the problem of low temperature estimation precision in related technologies. The method comprises the following steps: constructing an initial temperature estimation model based on a heat transfer network of to-be-tested equipment; in the operation state of the to-be-tested equipment, acquiring actual temperatures of the to-be-tested equipment in a plurality of preset temperature nodes in the heat transfer network, and estimating predicted temperatures of the preset temperature nodes based on the initial temperature estimation model; and adjusting the initial temperature estimation model through a preset error function based on the actual temperature and the predicted temperature of each preset temperature node to obtain a target temperature estimation model.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of motor control, and in particular to a temperature estimation model construction method and device, computer equipment and a storage medium. BACKGROUND

[0002] As the core execution unit of the driving system, the motor controller contains multiple key components that bear the functions of energy transmission and stabilization. These components will generate heat due to losses during operation, leading to temperature rise. Especially when running for a long time under continuous high current or harsh conditions, the component temperature may exceed the safety threshold, thereby triggering a series of serious risks. Therefore, it is crucial to monitor the real-time temperature of the key components inside the motor controller and take protective measures such as derating in a timely manner when the temperature exceeds the threshold, to ensure the reliability and safety of the motor controller.

[0003] In related technologies, a bus capacitor temperature online estimation method based on a heat transfer network model is proposed. This method measures the loss distribution of the relevant areas inside the motor controller under specific voltage, speed and torque conditions, constructs a "speed-torque-voltage" three-dimensional loss lookup table, and establishes a corresponding heat transfer network model. The model parameters in the heat transfer network model are calibrated through temperature rise tests of the motor controller under different conditions. Thus, in actual operation, the motor controller estimates the current temperature of the bus capacitor by combining the loss value obtained from the three-dimensional loss lookup table with the real-time temperature data collected.

[0004] However, this method mainly relies on offline experimental data under specific conditions to statically determine the model parameters during modeling, resulting in low temperature estimation accuracy of the heat transfer network model. SUMMARY

[0005] Therefore, a temperature estimation model method, device, computer equipment and storage medium are provided to solve the problem of low temperature estimation accuracy in related technologies.

[0006] In a first aspect, the present application provides a temperature estimation model construction method, which comprises: constructing an initial temperature estimation model based on the heat transfer network of the device to be tested; under the operating state of the device to be tested, collecting the actual temperatures of multiple preset temperature nodes in the heat transfer network of the device to be tested, and estimating the predicted temperatures of each preset temperature node based on the initial temperature estimation model; adjusting the initial temperature estimation model based on the actual and predicted temperatures of each preset temperature node through a preset error function to obtain a target temperature estimation model.

[0007] In one embodiment, the initial temperature estimation model is constructed based on the heat transfer network of the device under test, comprising: determining a temperature estimation model framework based on the heat transfer network of the device under test; determining temperature estimation parameter values according to the operating condition data of the device under test under different test operating conditions; generating the initial temperature estimation model according to the temperature estimation model framework and the temperature estimation parameter values.

[0008] In one embodiment, the initial temperature estimation model is adjusted to obtain a target temperature estimation model based on the actual temperature and the predicted temperature of each preset temperature node through a preset error function, comprising: obtaining the weight of each preset temperature node; calculating an error value based on the actual temperature and the predicted temperature of each preset temperature node within a preset running time period and the weight of each preset temperature node through the preset error function; adjusting the temperature estimation parameter values in the initial temperature estimation model according to the error value to obtain the target temperature estimation model.

[0009] In one embodiment, the preset error function is: ; wherein t0 is the start time of the preset running time period, t end is the end time of the preset running time period, Tact i is the actual temperature of the i-th preset temperature node, Test i is the predicted temperature of the i-th preset temperature node, ω i is the weight of the i-th preset temperature node, n is the total number of preset temperature nodes, and Err is the error value.

[0010] In one embodiment, the initial temperature estimation model is adjusted to obtain a target temperature estimation model according to the error value, comprising: searching in a preset solution space corresponding to the temperature estimation parameter values based on the error function with the objective of minimizing the error value to obtain target temperature estimation parameter values; adjusting the temperature estimation parameter values in the initial temperature estimation model according to the target temperature estimation parameter values to obtain the target temperature estimation model.

[0011] In a second aspect, the present application provides a temperature estimation method, which adopts the target temperature estimation model in the first aspect, and the method comprises: obtaining the actual operating parameters of the device under test; inputting the actual operation parameter into the target temperature estimation model to obtain an estimated temperature value of a target temperature node in the to-be-tested device.

[0012] In one embodiment, after obtaining the estimated temperature value of the target temperature node in the to-be-tested device, the method further comprises: In response to detecting that the to-be-tested device satisfies an initialization condition, initializing an initial temperature of each preset temperature node in the target temperature estimation model based on a current environment temperature.

[0013] In a third aspect, the present application provides a temperature estimation model construction device, which comprises: a first construction module configured to construct an initial temperature estimation model based on a heat transfer network of a to-be-tested device; a collection module configured to collect actual temperatures of a plurality of preset temperature nodes in the heat transfer network under an operation state of the to-be-tested device, and estimate predicted temperatures of the preset temperature nodes based on the initial temperature estimation model; a second construction module configured to adjust model parameters of the initial temperature estimation model based on the actual temperatures and the predicted temperatures of the preset temperature nodes through a preset error function to obtain a target temperature estimation model.

[0014] In a fourth aspect, the present application provides a temperature estimation device, which comprises: an acquisition module configured to acquire actual operation parameters of the to-be-tested device; a temperature estimation module configured to input the actual operation parameters into the target temperature estimation model to obtain an estimated temperature value of a target temperature node in the to-be-tested device.

[0015] In a fifth aspect, the present application provides a computer device, which comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the temperature estimation model construction method of the first aspect and the temperature estimation method of the second aspect when executing the computer program.

[0016] In a sixth aspect, the present application provides a computer readable storage medium, which stores a computer program, wherein the computer program is executable on a processor to implement the temperature estimation model construction method of the first aspect and the temperature estimation method of the second aspect.

[0017] The aforementioned temperature estimation model construction method, apparatus, computer equipment, and storage medium construct an initial temperature estimation model based on the heat transfer network of the device under test (DUT), establishing a theoretical thermal dynamic framework for the DUT. Then, the actual temperature is collected during operation and compared with the predicted temperature estimated by the model, effectively capturing the deviation between the theoretical model and the actual system. Finally, the model parameters are dynamically adjusted based on the actual and predicted temperatures, achieving online calibration and correction of the initial temperature estimation model. This results in a target temperature estimation model that accurately reflects the true thermal characteristics of the DUT, significantly improving the accuracy and robustness of temperature estimation under complex and changing operating conditions. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating a method for constructing a temperature estimation model in one embodiment; Figure 2 This is a schematic diagram of the heat transfer network in one embodiment; Figure 3 This is a flowchart illustrating a temperature estimation method in one embodiment; Figure 4 This is a structural block diagram of a temperature estimation model construction device in one embodiment; Figure 5 This is a structural block diagram of a temperature estimation device in one embodiment; Figure 6 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings. The specific operational methods in the method embodiments can also be applied to the device embodiments or system embodiments. It should be noted that in the description of this invention, "multiple" is understood as "at least two". "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing together, or B existing alone. A connected to B can represent: A and B directly connected, or A and B connected through C. Furthermore, in the description of this invention, terms such as "first" and "second" are used only for descriptive purposes and should not be construed as indicating or implying relative importance or order.

[0020] In this invention, the acquisition, transmission, storage, and use of data all comply with the requirements of relevant national laws and regulations.

[0021] Before introducing the temperature estimation model construction method provided by this invention, the technical background of this invention will be described in detail below for ease of understanding.

[0022] In related technologies, a method for online estimation of bus capacitor temperature based on a heat transfer network is proposed. This method, under specific voltage, speed, and torque conditions, pre-measures the loss distribution in relevant areas inside the motor controller, constructs a three-dimensional loss lookup table of "speed-torque-voltage," and establishes a corresponding heat transfer network model. The model parameters in the heat transfer network model are calibrated through temperature rise tests of the motor controller under different operating conditions. Therefore, in actual operation, the motor controller estimates the current temperature of the bus capacitor by combining real-time collected temperature data with the loss values ​​obtained from the three-dimensional loss lookup table.

[0023] However, this method relies primarily on offline, experimental data under specific operating conditions to statically determine model parameters during the modeling process. It fails to adequately consider the impact of thermal inertia (characterized by heat capacity) and heat transfer delay (characterized by thermal resistance) on the temperature response during actual dynamic operation, resulting in low estimation accuracy of the heat transfer network model. For example, when the load on the motor controller suddenly increases, the losses of the bus capacitor will immediately increase dramatically. However, due to thermal inertia and heat transfer delay, the actual temperature of the bus capacitor will not immediately spike to the corresponding steady-state temperature, but will instead experience a slow rise and eventually stabilize. This heat transfer network model ignores this delay, and its estimation results will instantaneously follow the changes in losses, failing to reflect the true temperature delay effect and leading to low estimation accuracy.

[0024] In view of this, the present invention provides a method, apparatus, computer device and storage medium for constructing a temperature estimation model, in order to solve the problem of low temperature estimation accuracy in related technologies.

[0025] The technical solution provided by the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0026] Figure 1 This is a flowchart illustrating a method for constructing a temperature estimation model in one embodiment. This process can be executed by a temperature estimation model construction device, which can be implemented in software, hardware, or a combination of both. Figure 1 As shown, the process includes the following steps: S101, Based on the heat transfer network of the device under test, construct an initial temperature estimation model; S102, under the operating state of the device under test, the actual temperature of multiple preset temperature nodes of the device under test in the heat transfer network is collected, and the predicted temperature of each preset temperature node is estimated based on the initial temperature estimation model. S103, by using a preset error function to adjust the initial temperature estimation model based on the actual and predicted temperatures of each preset temperature node, the target temperature estimation model is obtained.

[0027] The device under test can be any device that has thermal management requirements, whose internal temperature is critical but difficult to measure directly, and whose operating parameters are strongly correlated with thermal state, such as motor controllers, power battery packs, and on-board chargers. In this invention, a motor controller is used as an example for illustration.

[0028] For example, an initial temperature estimation model is constructed based on the heat transfer network of the device under test, wherein the heat transfer network includes preset temperature nodes inside the device under test and the heat transfer path of each preset temperature node.

[0029] Then, while the device under test is in operation, the actual temperatures of multiple preset temperature nodes in the heat transfer network are collected, and the predicted temperatures of each preset temperature node are estimated based on the initial temperature estimation model.

[0030] For example, in a software environment, an initial temperature estimation model is integrated into the simulation platform to simulate the actual operating state of the vehicle. During the simulation, a series of time-varying curves are input as driving conditions. These time-varying curves include, but are not limited to: electrical and load curves (bus voltage curve, motor speed curve, motor torque curve, etc.); and environmental and cooling curves (ambient temperature curve, coolant temperature curve, etc.). During the simulation, the actual temperature at each preset temperature node is recorded, and the initial temperature estimation model is run to output the predicted temperature at each preset temperature node.

[0031] Optionally, the initial temperature estimation model can also be deployed in the vehicle controller. During vehicle operation, the controller continuously acquires real-time operating parameters (such as losses and ambient temperature) of the device under test, enabling the initial temperature estimation model to estimate the predicted temperature of each preset temperature node based on these real-time operating parameters. Simultaneously, the actual temperature of each preset temperature node is collected through temperature sensors (such as thermocouples and thermistors) directly deployed at each preset temperature node.

[0032] Finally, by using a preset error function to adjust the initial temperature estimation model based on the actual and predicted temperatures of each preset temperature node, the target temperature estimation model is obtained.

[0033] In the above method, an initial temperature estimation model is constructed based on the heat transfer network of the device under test, establishing the theoretical thermal dynamic framework of the device under test. Then, the actual temperature is collected during operation and compared with the predicted temperature estimated by the model, effectively capturing the deviation between the theoretical model and the actual system. Finally, the model is dynamically adjusted by comparing the actual temperature and the predicted temperature, realizing online calibration and correction of the initial temperature estimation model. This results in a target temperature estimation model that accurately reflects the true thermal characteristics of the device under test, significantly improving the accuracy and robustness of temperature estimation under complex and changing operating conditions.

[0034] In one embodiment, it is illustrated by way of example that in S101, an initial temperature estimation model is constructed based on the heat transfer network of the device under test, including but not limited to: First, a temperature estimation model framework is determined based on the heat transfer network of the device under test. For example, a temperature estimation model framework is generated based on the heat transfer paths of each preset temperature node in the heat transfer network. The heat transfer paths are characterized by temperature estimation parameters, which include, but are not limited to, heat capacity parameters and thermal resistance parameters.

[0035] Then, based on the operating data of the device under test under different test conditions, the temperature estimation parameter values ​​are determined, and an initial temperature estimation model is generated according to the temperature estimation model framework and the temperature estimation parameter values. The operating data includes at least the wear and tear data of the device under test and the temperature sensor data.

[0036] For example, different test conditions can be obtained based on a fixed rotation speed (which fixes the convective heat dissipation conditions caused by the rotation speed, avoiding the coupling effect caused by the simultaneous change of multiple variables), different coolant temperatures (in order to identify the thermal resistance related to the cooling system), different ambient temperatures (in order to consider the thermal resistance of radiation and convection heat dissipation to the surrounding environment), different bus voltages, and different torques (torque directly determines the loss of the device under test; by changing the torque, a series of losses from low to high can be generated).

[0037] Under each test condition, the loss of the device under test is acquired, and temperature data at each preset temperature node is collected through temperature sensors. Based on the loss of the device under test and the temperature data at each preset temperature node acquired under all test conditions, the temperature estimation parameter values ​​are calculated by the system identification algorithm (such as least squares method, genetic algorithm, gradient descent method, etc., the specific algorithm depends on the situation and is not limited here). Thus, an initial temperature estimation model is generated according to the temperature estimation model framework and the temperature estimation parameter values.

[0038] In this invention, the DC bus capacitor and DC copper bus are important components of the motor controller. The DC copper bus is connected to the DC bus capacitor, thereby transferring electrical energy to the motor controller. The DC bus capacitor can suppress DC bus voltage fluctuations and provide electrical protection for power devices. However, when the motor controller operates at high current for a short period of time, the temperature of its internal DC copper bus rises rapidly, easily exceeding the normal operating temperature range of the DC copper bus, causing the plastic coating to soften. This reduces the hard creepage distance of the motor controller and leads to burn-out. When the motor controller operates at high current for a long period of time, the temperature of its internal DC bus capacitor will continue to rise, easily causing the capacitor to overheat and fail, resulting in control failure, damage to switching devices, and burn-out of the motor controller.

[0039] Therefore, the following section uses the estimation of the DC bus capacitor temperature and DC copper bus temperature inside the motor controller as an example to explain in detail the construction method of the initial temperature estimation model: First, multiple preset temperature nodes are obtained within the motor controller. These preset temperature nodes include at least a first temperature node representing the highest temperature of the DC copper bus inside the motor controller, a second temperature node representing the highest temperature of the DC bus capacitor inside the motor controller, a third temperature node representing the ambient temperature of the motor controller, and a fourth temperature node representing the temperature of the coolant flowing through the motor controller.

[0040] For example, by simulating temperature rise, a three-dimensional temperature field distribution cloud map of the motor controller is obtained, and based on this temperature field distribution cloud map, the temperature data of all nodes on the surface of the DC copper bus and DC bus capacitor are extracted; areas with temperature values ​​higher than a preset threshold (such as 110% of the average temperature of all temperature values, the specific threshold depends on the situation and is not limited here) and with a continuous distribution area larger than a preset area (such as 4 square millimeters, the specific area depends on the situation and is not limited here) are identified as hot spots, and the nodes with the highest temperature in the hot spots are identified as temperature measurement points.

[0041] Temperature sensors are placed at temperature measurement points, and the motor controller is controlled to operate under different speed and torque conditions to collect temperature data from the sensors. Analyzing this temperature data, if one of the multiple temperature measurement points corresponding to the DC bus capacitor consistently has a higher temperature value than the others, this first temperature measurement point is selected as the second temperature node representing the highest temperature of the DC bus capacitor. When the first temperature measurement point corresponding to the DC bus capacitor differs under different speed and torque conditions, these first temperature measurement points are all analyzed independently as second temperature nodes.

[0042] Similarly, analyzing the temperature values ​​of the positive and negative busbars of the DC busbar under different speed and torque conditions, when the temperature measurement point corresponding to the highest temperature value is located on either the positive or negative busbar under different speed and torque conditions, the temperature measurement point corresponding to the highest temperature value is selected as the first temperature node characterizing the highest temperature point of the DC busbar. When the temperature measurement point corresponding to the highest temperature value may be located on either the positive or negative busbar under different speed and torque conditions, the temperature measurement points corresponding to the highest temperature values ​​of both the positive and negative busbars need to be analyzed independently as the first temperature node.

[0043] In this invention, an example is taken where the temperature value of a first target temperature measurement point of the DC bus capacitor is always higher than other temperature values, and the temperature measurement point corresponding to the highest temperature value is located on the positive copper busbar.

[0044] Simultaneously, a third and fourth temperature node are determined. The third temperature node is the ambient temperature node, whose temperature value can be obtained in real time from temperature sensors located on the printed circuit board of the motor controller. Since the DC bus capacitor temperature and DC copper bus temperature will cause changes in the temperature of the fourth temperature node through heat conduction and heat radiation, the introduction of a third temperature node can reduce the mismatch between thermal resistance and thermal capacity parameters and actual thermal resistance and thermal capacity parameters caused by manufacturing and assembly errors in mass production, thereby improving the accuracy of temperature estimation.

[0045] The fourth temperature node is a virtual node because in a real motor system, the coolant itself is a flowing fluid whose temperature changes in the flow channel and is not a single, uniform temperature point. Furthermore, the temperature distribution of the entire flow field cannot be simulated in a simplified heat transfer network. Therefore, a virtual node is created to characterize the effective heat dissipation temperature of the coolant to the motor controller.

[0046] Thus, a heat transfer network is formed based on the first temperature node, the second temperature node, the third temperature node, and the fourth temperature node.

[0047] Figure 2 This is a schematic diagram of the heat transfer network in one embodiment, such as... Figure 2 As shown, it includes a first temperature node T1, a second temperature node T2, a third temperature node T3, and a fourth temperature node T4. The heat source for T1 is the heat generated by the DC copper busbar loss P1. T1 is connected to the first thermal capacitor C1 and connected to the first thermal resistor R. 12 Connected to T2, through the second thermal resistance R 14 Connected to T4; the heat source of T2 is the heat generated by the DC bus capacitor loss P2. T2 is also connected to the second thermal capacitor C2 and connected to the third thermal resistor R. 23 It is connected to T3; T3 is represented by a known temperature through a voltage source U, and is also connected to the third thermal capacity C3, and through the fourth thermal resistance R. 34 It is connected to T4; T4 is also connected to the fourth heat capacity C4.

[0048] Then, based on the heat transfer paths of each preset temperature node in the heat transfer network, the framework of the temperature estimation model is determined.

[0049] like Figure 2 As shown, in this heat transfer network, by analogy with Kirchhoff's current law for each preset temperature node, i.e., the sum of all currents flowing into a node is zero, the first heat transfer path of T1 is obtained as follows: The second heat transfer path of T2 is: The third heat transfer path of T3 is: The fourth heat transfer path of T4 is: Therefore, based on the first, second, third, and fourth heat transfer paths, the framework of the temperature estimation model is determined as follows: .

[0050] Finally, based on the DC copper bus loss, DC bus capacitance loss, and temperature data of the first, second, third, and fourth temperature nodes under different test conditions, the values ​​of each thermal resistance parameter (R12, R14, R23, R34) and each thermal capacity parameter (C1, C2, C3, C4) are calculated. Based on the temperature estimation model framework, the values ​​of each thermal resistance parameter, and the values ​​of each thermal capacity parameter, an initial temperature estimation model is generated.

[0051] The methods for obtaining DC bus capacitor loss include, but are not limited to: determining DC bus capacitor loss by looking up a table based on the operating current information of the motor controller.

[0052] For example, a current loss table can be pre-established through simulation. This table contains the mapping relationship between the average value of AC current and the DC bus capacitor loss. Thus, under different operating conditions, the corresponding DC bus capacitor loss can be directly found from the current loss table based on the average value of the three-phase AC current of the motor controller at the current moment.

[0053] Alternatively, the effective value of the ripple current I1 flowing through the DC bus capacitor can be determined through simulation, and its equivalent series resistance (ESR) can be obtained from the DC bus capacitor's datasheet. Then, the DC bus capacitor loss = I1 can be calculated based on I1 and ESR. 2 ×ESR. The specific method for obtaining DC bus capacitance loss depends on the situation and is not limited here.

[0054] Methods for obtaining DC copper busbar losses include, but are not limited to: using a high-precision millivoltmeter to directly measure the voltage drop ΔU across the DC copper busbar, while simultaneously measuring the current I2 flowing through the DC copper busbar, and calculating the DC copper busbar loss = ΔU × I2 based on ΔU and I2.

[0055] Optionally, the DC busbar at a reference temperature T can also be obtained. b (e.g., 25 degrees Celsius; the specific temperature depends on the situation and is not limited here) The reference impedance R measured at this temperature. b And based on R b The temperature compensation method is used to obtain the actual impedance of the DC copper busbar, and based on the actual impedance and the DC current flowing through the DC copper busbar, the DC copper busbar loss is calculated. The specific method for obtaining the DC copper busbar loss depends on the situation and is not limited here. Specifically: Obtain the estimated temperature value T of the first temperature node output by the initial temperature estimation model at the current moment.act And based on T act With T b The temperature difference between them (T) act -T b ), for R b Compensation is performed to obtain the actual impedance of the DC busbar. For example, the impedance temperature coefficient R of the DC busbar is obtained. TF , where R TF Determined based on the temperature coefficient of resistance of DC copper busbar material, and based on (T act -T b ), R TF and R b The actual impedance of the DC copper busbar is calculated to be (1 + (T)). act -T b )×R TF )×R b .

[0056] Simultaneously, the DC current I2 flowing through the DC copper busbar is obtained, and based on I2 and the actual impedance, the DC copper busbar loss is calculated as: I2 × (1 + (T)) / (T) act -T b )×R TF )×R b .

[0057] For example, the estimated temperature value of the first temperature node output by the initial temperature estimation model at time (t-1) is T. （t-1） Obtain the DC current I flowing through the DC busbar at time t. t Then based on T （t-1） and I t The DC copper busbar loss at time t can be calculated, and by inputting the DC copper busbar loss into the initial temperature estimation model, the estimated temperature value of the first temperature node at time t can be obtained.

[0058] Using the above method, the key temperature nodes of the heat transfer network of the device under test are accurately selected. Based on the heat transfer network formed by the key temperature nodes, the temperature estimation model framework defined by the temperature estimation parameters is determined. Then, the temperature estimation parameter values ​​are calibrated using steady-state temperature rise tests. Thus, an initial temperature estimation model with clear physical meaning and accurate calibration under steady-state conditions is established, laying a reliable theoretical foundation for the online optimization of the model under subsequent dynamic operating conditions.

[0059] In one embodiment, exemplarily illustrated, S103 adjusts the initial temperature estimation model based on the actual and predicted temperatures of each preset temperature node using a preset error function to obtain a target temperature estimation model, including but not limited to: Obtain the weights of each preset temperature node. The weights are assigned based on the importance of the temperature accuracy of each preset temperature node. Temperature nodes with higher importance can be assigned higher weights, and temperature nodes with lower importance can be assigned lower weights. The specific assignment method depends on the situation and is not limited here.

[0060] For example, in a motor controller, the DC bus capacitor is the most easily damaged component and has the highest importance, so the weight of the second temperature node is 1.5. The DC copper busbar is the second most easily damaged component besides the DC bus capacitor and has the next highest importance, so the weight of the first temperature node is 1. The weights of all other temperature nodes are less than 1.

[0061] Then, the error value is calculated based on the actual temperature and predicted temperature of each preset temperature node within a preset operating time period (the specific time period depends on the situation and is not limited here) and the weight of each preset temperature node through a preset error function. Finally, the temperature estimation parameters in the initial temperature estimation model are adjusted based on the error value to obtain the target temperature estimation model.

[0062] By introducing the above method and supervising learning with a weighting mechanism, the initial temperature estimation model is accurately calibrated, thereby significantly improving the accuracy and overall reliability of the temperature estimation model at key temperature nodes.

[0063] In one embodiment, the preset error function is illustrated by way of example: (1); In formula (1), t0 is the start time of the preset running time period, t end Tact sets the end time of the preset running time period. i For the actual temperature of the i-th preset temperature node, Test i ω is the predicted temperature of the i-th preset temperature node. i Let be the weight of the i-th preset temperature node, n be the total number of preset temperature nodes, and Err be the error value.

[0064] Using the above method, based on a preset error function, the overall fitting deviation of the model to all preset temperature nodes within a preset running time period is comprehensively evaluated, thereby guiding the adjustment direction of the model parameters, and finally training a high-precision temperature estimation model that performs stably in key locations and continuous operation.

[0065] In one embodiment, it is exemplarily illustrated that adjusting the temperature estimation parameter values ​​in the initial temperature estimation model based on the error value yields the target temperature estimation model, including but not limited to: With minimizing Err as the objective, the target temperature estimation parameter value is obtained by searching within the preset solution space corresponding to the temperature estimation parameter value based on the error function. The temperature estimation parameter value in the initial temperature estimation model is then adjusted based on the target temperature estimation parameter value to obtain the target temperature estimation model.

[0066] For example, for each thermal resistance parameter and each heat capacity parameter value in the initial temperature estimation model, an error range for parameter optimization is set (determined based on empirical values), resulting in a preset solution space for each parameter value. Then, using a random sampling method, multiple sets of parameter value combinations are generated based on the preset solution space for each parameter value. With minimizing Err as the objective, multiple sets of parameter value combinations are evaluated and optimized based on an error function to determine the target thermal resistance parameter value and target heat capacity parameter value combination that minimizes Err. This target thermal resistance parameter value and target heat capacity parameter value combination are then updated to the initial temperature estimation model, resulting in the target temperature estimation model.

[0067] By using the above method, with the goal of minimizing the error value, an intelligent search is performed in the preset solution space corresponding to the temperature estimation parameter value to accurately identify the optimal target temperature estimation parameter value. This achieves the calibration of key parameters of the initial temperature estimation model and significantly improves the estimation accuracy and adaptability of the model.

[0068] In this invention, by constructing a target temperature estimation model, the temperature value of the target temperature node in the device under test can be estimated in real time. When the temperature value of the target temperature node exceeds the normal operating temperature range, the device under test is automatically controlled to operate at a reduced rate, thereby preventing thermal failure of the device under test.

[0069] Figure 3 This is a flowchart illustrating a temperature estimation method in one embodiment. The method employs... Figure 1 The target temperature estimation model in the process can be executed by a temperature estimation device, which can be implemented through software, hardware, or a combination of both. For example... Figure 3 As shown, the process includes the following steps: S301, Obtain the actual operating parameters of the device under test; S302, input the actual operating parameters into the target temperature estimation model to obtain the estimated temperature value of the target temperature node in the device under test.

[0070] The actual operating parameters include at least the wear and tear of the device under test and the ambient temperature.

[0071] For example, the first loss of the DC copper bus in the motor controller (which is obtained in the same way as the DC copper bus loss described above, and will not be described again here), the second loss of the DC bus capacitor (which is obtained in the same way as the DC bus capacitor loss described above, and will not be described again here), and the ambient temperature (which is obtained in the same way as the temperature of the third temperature node described above, and will not be described again here) are obtained. The first loss, the second loss, and the ambient temperature are then input into the target temperature estimation model to obtain the estimated temperature value of the first temperature node and the estimated temperature value of the second temperature node.

[0072] The above temperature estimation method obtains the actual operating parameters of the device under test and inputs them into the target temperature estimation model that has been dynamically calibrated. This enables online and accurate estimation of the actual operating temperature of the target temperature node inside the device under test. Thus, without increasing hardware costs, it provides key status information for the overheat protection and life prediction of the device under test, effectively improving the reliability and safety of the device under test.

[0073] In one embodiment, exemplarily illustrated, after obtaining the estimated temperature value of the target temperature node in the device under test in step S302, the method further includes: In response to the detection that the device under test meets the initialization conditions, the initial temperature of each preset temperature node in the target temperature estimation model is initialized based on the current ambient temperature.

[0074] For example, in response to detecting that the motor controller is powered down, or that the cooling time of the motor controller exceeds a preset time (ensuring that the motor controller has undergone a preset cooling time, and the temperature of its internal components has dropped to a state sufficiently close to the ambient temperature; the specific time depends on the situation and is not limited here), the temperature of each preset temperature node in the target temperature estimation model is initialized to the current ambient temperature (consistent with the temperature acquisition method of the third temperature node mentioned above, and will not be described again here), or the temperature deviation of each preset temperature node is acquired (determined based on empirical values), and the temperature of each preset temperature node is initialized to the sum of the corresponding temperature deviation and the current ambient temperature. The specific initialization method depends on the situation and is not limited here.

[0075] By using the above method, when the device under test is detected to meet the initialization conditions, the internal state of the model is automatically refreshed using the current ambient temperature, thereby effectively eliminating the cumulative error of the model's historical thermal state on the new round of temperature estimation, and ensuring the continuous accuracy and adaptability of the temperature estimation model in the long term and under different usage scenarios.

[0076] It should be understood that, although Figure 1 and 3The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figure 1 and 3 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0077] In one embodiment, such as Figure 4 As shown, a temperature estimation model construction device is provided, including: a first construction module 401, a data acquisition module 402, and a second construction module 403, wherein: The first construction module 401 is used to construct an initial temperature estimation model based on the heat transfer network of the device under test. The acquisition module 402 is used to acquire the actual temperature of multiple preset temperature nodes in the heat transfer network when the device under test is in operation, and to estimate the predicted temperature of each preset temperature node based on the initial temperature estimation model. The second construction module 403 is used to adjust the model parameters of the initial temperature estimation model based on the actual temperature and predicted temperature of each preset temperature node through a preset error function, so as to obtain the target temperature estimation model.

[0078] In one embodiment, the first building module 401 is used for: Based on the heat transfer network of the device under test, a temperature estimation model framework is determined. Based on the operating data of the device under test under different test conditions, the temperature estimation parameter values ​​are determined. Based on the temperature estimation model framework and temperature estimation parameter values, an initial temperature estimation model is generated.

[0079] In one embodiment, the second building module 403 is used for: Obtain the weights of each preset temperature node; The error value is calculated by using a preset error function based on the actual and predicted temperatures of each preset temperature node within a preset operating time period, as well as the weight of each preset temperature node. Adjust the temperature estimation parameter values ​​in the initial temperature estimation model based on the error value to obtain the target temperature estimation model.

[0080] In one embodiment, the preset error function is: ; Where t0 is the start time of the preset running time period, t end Tact sets the end time of the preset running time period. i For the actual temperature of the i-th preset temperature node, Test i ω is the predicted temperature of the i-th preset temperature node. i Let be the weight of the i-th preset temperature node, n be the total number of preset temperature nodes, and Err be the error value.

[0081] In one embodiment, the second building module 403 is further configured to: With the goal of minimizing the error value, the target temperature estimation parameter value is obtained by searching within the preset solution space corresponding to the temperature estimation parameter value based on the error function. The target temperature estimation model is obtained by adjusting the temperature estimation parameter values ​​in the initial temperature estimation model based on the target temperature estimation parameter values.

[0082] Specific limitations regarding the temperature estimation model construction device can be found in the limitations of the temperature estimation model construction method described above, and will not be repeated here. Each module in the aforementioned temperature estimation model construction device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0083] In one embodiment, such as Figure 5 As shown, a temperature estimation device is provided, including: an acquisition module 501 and a temperature estimation module 502, wherein: The acquisition module 501 is used to acquire the actual operating parameters of the device under test. The temperature estimation module 502 is used to input the actual operating parameters into the target temperature estimation model to obtain the estimated temperature value of the target temperature node in the device under test.

[0084] In one embodiment, the temperature estimation module 502 is further configured to: In response to the detection that the device under test meets the initialization conditions, the initial temperature of each preset temperature node in the target temperature estimation model is initialized based on the current ambient temperature.

[0085] Specific limitations regarding the temperature estimation device can be found in the limitations of the temperature estimation method described above, and will not be repeated here. Each module in the aforementioned temperature estimation device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in hardware or independently of the processor in a computer device, or stored in software in the memory of a computer device, so that the processor can call and execute the corresponding operations of each module.

[0086] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 6 As shown. The computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The database stores temperature estimation model construction data and temperature estimation data. The network interface communicates with external terminals via a network connection. When the processor executes the computer program, it implements a temperature estimation model construction method and a temperature estimation method. The display screen can be an LCD screen or an e-ink display screen. The input device can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device casing, or an external keyboard, touchpad, or mouse.

[0087] Those skilled in the art will understand that Figure 6 The structure shown is merely a block diagram of a portion of the structure related to the present invention and does not constitute a limitation on the computer device to which the present invention is applied. A specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0088] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to perform the following steps: An initial temperature estimation model is constructed based on the heat transfer network of the device under test. While the device under test is in operation, the actual temperature of multiple preset temperature nodes in the heat transfer network of the device under test is collected, and the predicted temperature of each preset temperature node is estimated based on the initial temperature estimation model. By adjusting the initial temperature estimation model based on the actual and predicted temperatures of each preset temperature node using a preset error function, the target temperature estimation model is obtained.

[0089] In one embodiment, the processor, when executing a computer program, also performs the following steps: Based on the heat transfer network of the device under test, a temperature estimation model framework is determined. Based on the operating data of the device under test under different test conditions, the temperature estimation parameter values ​​are determined. Based on the temperature estimation model framework and temperature estimation parameter values, an initial temperature estimation model is generated.

[0090] In one embodiment, the processor, when executing a computer program, also performs the following steps: Obtain the weights of each preset temperature node; The error value is calculated by using a preset error function based on the actual and predicted temperatures of each preset temperature node within a preset operating time period, as well as the weight of each preset temperature node. Adjust the temperature estimation parameter values ​​in the initial temperature estimation model based on the error value to obtain the target temperature estimation model.

[0091] In one embodiment, the processor, when executing a computer program, also performs the following steps: With the goal of minimizing the error value, the target temperature estimation parameter value is obtained by searching within the preset solution space corresponding to the temperature estimation parameter value based on the error function. The target temperature estimation model is obtained by adjusting the temperature estimation parameter values ​​in the initial temperature estimation model based on the target temperature estimation parameter values.

[0092] In one embodiment, the processor, when executing a computer program, also performs the following steps: Obtain the actual operating parameters of the device under test; By inputting the actual operating parameters into the target temperature estimation model, the estimated temperature value of the target temperature node in the device under test is obtained.

[0093] In one embodiment, the processor, when executing a computer program, also performs the following steps: In response to the detection that the device under test meets the initialization conditions, the initial temperature of each preset temperature node in the target temperature estimation model is initialized based on the current ambient temperature.

[0094] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor: An initial temperature estimation model is constructed based on the heat transfer network of the device under test. While the device under test is in operation, the actual temperature of multiple preset temperature nodes in the heat transfer network of the device under test is collected, and the predicted temperature of each preset temperature node is estimated based on the initial temperature estimation model. By adjusting the initial temperature estimation model based on the actual and predicted temperatures of each preset temperature node using a preset error function, the target temperature estimation model is obtained.

[0095] In one embodiment, when the computer program is executed by a processor, it also performs the following steps: Based on the heat transfer network of the device under test, a temperature estimation model framework is determined. Based on the operating data of the device under test under different test conditions, the temperature estimation parameter values ​​are determined. Based on the temperature estimation model framework and temperature estimation parameter values, an initial temperature estimation model is generated.

[0096] In one embodiment, when the computer program is executed by a processor, it also performs the following steps: Obtain the weights of each preset temperature node; The error value is calculated by using a preset error function based on the actual and predicted temperatures of each preset temperature node within a preset operating time period, as well as the weight of each preset temperature node. Adjust the temperature estimation parameter values ​​in the initial temperature estimation model based on the error value to obtain the target temperature estimation model.

[0097] In one embodiment, when the computer program is executed by a processor, it also performs the following steps: With the goal of minimizing the error value, the target temperature estimation parameter value is obtained by searching within the preset solution space corresponding to the temperature estimation parameter value based on the error function. The target temperature estimation model is obtained by adjusting the temperature estimation parameter values ​​in the initial temperature estimation model based on the target temperature estimation parameter values.

[0098] In one embodiment, when the computer program is executed by a processor, it also performs the following steps: Obtain the actual operating parameters of the device under test; By inputting the actual operating parameters into the target temperature estimation model, the estimated temperature value of the target temperature node in the device under test is obtained.

[0099] In one embodiment, when the computer program is executed by a processor, it also performs the following steps: In response to the detection that the device under test meets the initialization conditions, the initial temperature of each preset temperature node in the target temperature estimation model is initialized based on the current ambient temperature.

[0100] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided by this invention can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0101] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0102] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.

Claims

1. A method for constructing a temperature estimation model, characterized in that, The method includes: An initial temperature estimation model is constructed based on the heat transfer network of the device under test. While the device under test is in operation, the actual temperatures of multiple preset temperature nodes of the device under test in the heat transfer network are collected, and the predicted temperatures of each preset temperature node are estimated based on the initial temperature estimation model. The initial temperature estimation model is adjusted based on the actual and predicted temperatures of each preset temperature node using a preset error function to obtain the target temperature estimation model.

2. The method according to claim 1, characterized in that, The initial temperature estimation model is constructed based on the heat transfer network of the device under test, including: Based on the heat transfer network of the device under test, a temperature estimation model framework is determined. Based on the operating data of the device under test under different test conditions, the temperature estimation parameter values ​​are determined. The initial temperature estimation model is generated based on the temperature estimation model framework and the temperature estimation parameter values.

3. The method according to claim 1, characterized in that, The step of adjusting the initial temperature estimation model based on the actual and predicted temperatures of each preset temperature node using a preset error function to obtain the target temperature estimation model includes: Obtain the weights of each of the preset temperature nodes; The error value is calculated using the preset error function based on the actual and predicted temperatures of each preset temperature node within a preset operating time period, as well as the weights of each preset temperature node. The temperature estimation parameter values ​​in the initial temperature estimation model are adjusted based on the error value to obtain the target temperature estimation model.

4. The method according to claim 3, characterized in that, The preset error function is: ; Where t0 is the start time of the preset running time period, t end The end time of the preset running time period, Tact i For the actual temperature of the i-th preset temperature node, Test i ω is the predicted temperature of the i-th preset temperature node. i Let be the weight of the i-th preset temperature node, n be the total number of preset temperature nodes, and Err be the error value.

5. The method according to claim 3, characterized in that, The step of adjusting the temperature estimation parameter values ​​in the initial temperature estimation model based on the error value to obtain the target temperature estimation model includes: With the goal of minimizing the error value, the target temperature estimation parameter value is obtained by searching within a preset solution space corresponding to the temperature estimation parameter value based on the error function. The target temperature estimation model is obtained by adjusting the temperature estimation parameter values ​​in the initial temperature estimation model based on the target temperature estimation parameter values.

6. A temperature estimation method, employing the target temperature estimation model according to any one of claims 1-5, characterized in that, The method includes: Obtain the actual operating parameters of the device under test; The actual operating parameters are input into the target temperature estimation model to obtain the estimated temperature value of the target temperature node in the device under test.

7. The method according to claim 6, characterized in that, After obtaining the estimated temperature value of the target temperature node in the device under test, the method further includes: In response to the detection that the device under test meets the initialization conditions, the initial temperature of each preset temperature node in the target temperature estimation model is initialized based on the current ambient temperature.

8. A device for constructing a temperature estimation model, characterized in that, The device includes: The first building module is used to construct an initial temperature estimation model based on the heat transfer network of the device under test. The acquisition module is used to acquire the actual temperature of multiple preset temperature nodes in the heat transfer network under the operating state of the device under test, and to estimate the predicted temperature of each preset temperature node based on the initial temperature estimation model. The second construction module is used to adjust the model parameters of the initial temperature estimation model based on the actual temperature and predicted temperature of each preset temperature node using a preset error function, so as to obtain the target temperature estimation model.

9. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method of any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 7.

Images