Temperature control method, robot, electronic device and medium

Abstract

The present application provides a temperature control method, robot, electronic device, and medium. The temperature control method includes: optimizing a control parameter sequence of a heat sink device in a future time period based on a cost function to obtain an optimized control parameter sequence; obtaining the real-time temperature of a heat source device to obtain a real-time temperature parameter sequence; determining the real-time control parameter of the heat sink device based on the real-time temperature parameter sequence and a preset temperature parameter threshold sequence; determining a target control parameter based on the optimized control parameter sequence and the real-time control parameter; and controlling the operation of the heat sink device based on the target control parameter to achieve temperature control of the heat source device. The technical solution of the present application can achieve early control of the temperature of the heat source device and improve the timeliness of temperature control to a certain extent.

Description

Technical Field

[0001] The present application relates to the field of temperature control technology, and in particular to a temperature control method, a robot, an electronic device, and a medium. Background Art

[0002] Thermal management equipment, such as robots, typically contains multiple high-power density or high-power consumption components, such as motors, power supply voltage regulators, high-performance computing units, and lidar. Robots are prone to localized overheating during extended operation or under high-load conditions. This can lead to device performance degradation or damage, or even operational failure. Therefore, effective methods are needed to achieve precise thermal management of robots. Current temperature control methods suffer from poor control effectiveness, including lags in the temperature control process. Summary of the Invention

[0003] In view of this, the embodiments of the present application provide a temperature control method, a robot, an electronic device, and a medium, which can achieve advance control of the temperature of a heat source device and improve the timeliness of temperature control to a certain extent.

[0004] In a first aspect, an embodiment of the present application provides a temperature control method, comprising: optimizing a control quantity parameter sequence of a heat dissipation device in a future time period based on a cost function to obtain an optimized control quantity parameter sequence; acquiring a real-time temperature of a heat source device to obtain a real-time temperature parameter sequence, and determining the real-time control quantity parameters of the heat dissipation device according to the real-time temperature parameter sequence and a preset temperature parameter threshold sequence; determining a target control quantity parameter based on the optimized control quantity parameter sequence and the real-time control quantity parameter; and controlling the operation of the heat dissipation device based on the target control quantity parameter to achieve temperature control of the heat source device.

[0005] In the second aspect, an embodiment of the present application provides a temperature control device, including: an optimization module, used to optimize the control quantity parameter sequence of the heat dissipation device in the future time period based on the cost function, and obtain the optimized control quantity parameter sequence; an acquisition module, used to obtain the real-time temperature of the heat source device, obtain the real-time temperature parameter sequence, and determine the real-time control quantity parameters of the heat dissipation device according to the real-time temperature parameter sequence and the preset temperature parameter threshold sequence; a determination module, used to determine the target control quantity parameters based on the optimized control quantity parameter sequence and the real-time control quantity parameters; a control module, used to control the operation of the heat dissipation device based on the target control quantity parameters, so as to achieve control of the temperature of the heat source device.

[0006] In a third aspect, an embodiment of the present application provides a robot comprising a control module, the control module being configured to execute the temperature control method described in the first aspect above.

[0007] In a fourth aspect, an embodiment of the present application provides an electronic device, comprising: a processor; and a memory for storing processor-executable instructions, wherein the processor is used to execute the temperature control method described in the first aspect above.

[0008] In a fifth aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the storage medium stores a computer program, and the computer program is used to execute the temperature control method described in the first aspect above.

[0009] In a sixth aspect, an embodiment of the present application provides a computer program product, which includes a computer program. When the computer program is executed by a processor of a computer device, the computer device is able to execute the temperature control method described in the first aspect above.

[0010] In a seventh aspect, an embodiment of the present application provides a chip comprising: a processor; and a memory for storing processor-executable instructions, wherein the processor is used to execute the temperature control method described in the first aspect above.

[0011] Embodiments of the present application provide a temperature control method, robot, electronic device, and medium. By predicting control parameters of a heat sink, the control parameters of the heat sink can be predicted or optimized in advance for a future time period. Since the control parameters can affect the temperature parameters of the heat source, the optimized control parameters can be used to control the temperature of the heat source to minimize exceeding a temperature parameter threshold during the future time period. This allows for proactive control of the heat source temperature, enabling early response to temperature rises, reducing the probability of emergency braking and frequent heat sink activation, thereby reducing system energy consumption. Furthermore, by proactively controlling the heat source temperature, the temperature parameter threshold can be set more accurately, eliminating the need for empirical estimation. Furthermore, by determining the real-time control parameters of the heat sink based on the real-time temperature of the heat source, determining target control parameters based on the optimized and real-time control parameters for the future time period, and controlling the operation of the heat sink based on the target control parameters, the timely temperature control can be improved based on proactive temperature control of the heat source, thereby addressing sudden temperature changes. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 Shown is a schematic diagram of the system architecture of a temperature control system provided by an exemplary embodiment of the present application.

[0013] Figure 2 Shown is a flow chart of a temperature control method provided by an exemplary embodiment of the present application.

[0014] Figure 3Shown is a flow chart of a temperature control method provided by another exemplary embodiment of the present application.

[0015] Figure 4 Shown is a schematic structural diagram of a temperature control device provided by an exemplary embodiment of the present application.

[0016] Figure 5 FIG2 is a block diagram of an electronic device for executing a temperature control method provided by an exemplary embodiment of the present application. DETAILED DESCRIPTION

[0017] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0018] Application Overview

[0019] Thermal management devices, such as robots, typically contain multiple high-power density or high-power consumption components. Robots are prone to localized overheating during extended operation or under high-load conditions. Currently, temperature sensors and heat dissipation devices, such as fans, are typically installed near high-power consumption components (heat source components). These temperature sensors monitor the temperature of these components. When the temperature exceeds a set safety threshold, the fan speed is increased to control the temperature to a certain extent, preventing overheating and potentially damaging the components. However, this temperature control method only increases the fan speed when the actual temperature of the high-power component exceeds the safety threshold. This introduces a certain lag, making it prone to untimely temperature control. Furthermore, this temperature control method requires high technical expertise, requiring technicians to set a reasonable safety threshold based on experience. Furthermore, this lag makes it prone to emergency braking in some scenarios, such as significantly reducing the load on the high-power component, leading to unstable operation of the high-power component. Furthermore, in some scenarios, the fan is prone to frequent operation, such as increasing its speed, which increases system energy consumption.

[0020] In summary, the current temperature control methods have poor control effects, such as the problem of lag in the temperature control process and the problem of high experience requirements for technicians.

[0021] To address the above technical issues, embodiments of the present application provide a temperature control method that, by predicting control parameters of a heat sink, can predict or optimize the control parameters of the heat sink for a future time period in advance. Since the control parameters can affect the temperature parameters of the heat source device, the optimized control parameters can be used to control the temperature of the heat source device to minimize exceeding a temperature parameter threshold during the future time period. This enables proactive control of the heat source device's temperature, enabling early response to temperature rises, reducing the probability of emergency braking and frequent heat sink activation, thereby reducing system energy consumption. Furthermore, by proactively controlling the heat source device's temperature, the temperature parameter threshold can be set more accurately, eliminating the need for empirical estimation. Furthermore, by determining the real-time control parameters of the heat sink based on the real-time temperature of the heat source device, determining target control parameters based on the optimized control parameters and the real-time control parameters for the future time period, and controlling the operation of the heat sink based on the target control parameters, the timely temperature control can be improved based on proactive temperature control of the heat source device, thereby addressing sudden temperature changes.

[0022] Exemplary Systems

[0023] Figure 1 The following is a schematic diagram of the system architecture of a temperature control system provided by an exemplary embodiment of the present application. Figure 1 As shown, the temperature control system 100 may include a controller 110, a heat source device 120, and a heat sink 130. The heat sink 130 is used to dissipate heat from the heat source device 120. The controller 110 is used to determine a control quantity parameter of the heat sink 130, which can be used to represent the speed of the motor of the heat sink 130.

[0024] Specifically, the controller 110 can optimize the control parameter sequence of the heat sink 130 in the future time period based on a cost function to obtain an optimized control parameter sequence. The cost function can be constructed based on the difference between the temperature parameter sequence of the heat source device 120 in the future time period and the preset temperature parameter threshold sequence, as well as the control parameter sequence of the heat sink 130. The temperature parameter sequence includes the temperature parameters of the heat source device 120 at multiple times in the future time period. The values ​​of the control parameter sequence affect the values ​​of the temperature parameter sequence, that is, the control parameter of the heat sink 130 affects the temperature of the heat source device 120.

[0025] Furthermore, the controller 110 may obtain the real-time temperature of the heat source device 120 , obtain a real-time temperature parameter sequence, and determine the real-time control quantity parameters of the heat dissipation device 130 according to the real-time temperature parameter sequence and a preset temperature parameter threshold sequence.

[0026] Furthermore, the controller 110 may determine target control quantity parameters based on the optimized control quantity parameter sequence and the real-time control quantity parameters, and control the operation of the heat dissipation device 130 based on the target control quantity parameters to achieve temperature control of the heat source device 120 .

[0027] In one example, the controller 110 , the heat source device 120 , and the heat dissipation device 130 may be provided on a first computing device.

[0028] Alternatively, in another example, the controller 110 may be deployed on a first computing device, and the heat source device 120 and the heat sink 130 may be provided on a second computing device. The first computing device and the second computing device may be communicatively connected. For example, the controller 110 in the first computing device may optimize the control quantity parameter sequence of the heat sink 130 in a future time period based on a cost function to obtain an optimized control quantity parameter sequence. The second computing device may obtain the real-time temperature of the heat source device 120, obtain a real-time temperature parameter sequence, and send the real-time temperature parameter sequence to the first computing device. The controller 110 in the first computing device may determine the real-time control quantity parameters of the heat sink 130 based on the real-time temperature parameter sequence and the preset temperature parameter threshold sequence, determine the target control quantity parameters based on the optimized control quantity parameter sequence and the real-time control quantity parameters, and send the target control quantity parameters to the second computing device. The second computing device may control the operation of the heat sink 130 based on the target control quantity parameters to achieve control of the temperature of the heat source device 120.

[0029] In one example, either the first computing device or the second computing device may be a server or an electronic device. The server may be a physical server, a cloud server, a virtual server, or another type of server. The electronic device may include a computer, a mobile phone, a tablet, or another mobile device. The mobile device may include a robot or a car. The robot may be a humanoid robot, a wheeled robot, a tracked robot, or another type of robot.

[0030] It should be understood that the above application scenario examples are only provided to facilitate understanding of the spirit and principles of the present application, and the embodiments of the present application are not limited thereto. On the contrary, the embodiments of the present application can be applied to any applicable scenario.

[0031] Exemplary Methods

[0032] Figure 2 Shown is a flow chart of a temperature control method provided by an exemplary embodiment of the present application. Figure 2 The method can be Figure 1 The controller 110 in the embodiment is executed. Figure 2 As shown, the temperature control method may include the following contents.

[0033] 210: Optimizing a control parameter sequence of the heat dissipation device in a future time period based on the cost function to obtain an optimized control parameter sequence.

[0034] Specifically, the heat dissipation device is used to dissipate heat from the heat source device. The cost function is constructed based on the difference between the temperature parameter sequence of the heat source device in the future time period and the preset temperature parameter threshold sequence, and the control parameter sequence of the heat dissipation device. The temperature parameter sequence includes the temperature parameters of the heat source device at multiple moments in the future time period. The value of the control parameter sequence affects the value of the temperature parameter sequence.

[0035] In one example, the heat source device may include at least one of the following: a metal-oxide-semiconductor field-effect transistor (MOSFET), a coil, a linear regulator, a power module, a high-frequency transformer, a power amplifier, a heating rod, or other devices that generate heat dissipation.

[0036] In one example, the heat dissipation device may include at least one of the following: an air-cooled heat dissipation device, a liquid-cooled heat dissipation device, or another type of heat dissipation device. The air-cooled heat dissipation device may include a fan, and the liquid-cooled heat dissipation device may include a pump.

[0037] In one example, a temperature parameter may represent the temperature of a heat source component, and a temperature parameter sequence may include temperature parameters corresponding to the heat source component at multiple moments in a future time period. A preset temperature parameter threshold sequence may include temperature parameter thresholds corresponding to the heat source component. The temperature parameter thresholds may be set in advance based on the physical properties of the heat source component. The number of temperature parameter thresholds included in the preset temperature parameter threshold sequence may be consistent with the number of heat source components.

[0038] In one example, a control parameter sequence may include multiple control parameters, each of which may correspond to multiple moments in a future time period. The control parameters may be used to control the speed of a motor in a heat sink. For example, a larger value of the control parameter increases the speed of the motor in the heat sink, which in turn reduces the temperature of the heat source. Therefore, the value of the control parameter affects the value of the temperature parameter; that is, the value of the control parameter sequence affects the value of the temperature parameter sequence.

[0039] In one example, the control parameter can be called a duty cycle, which represents the proportion of time a signal is in an active state (e.g., a high level, high or maximum speed, or an on-state) within a cycle. This parameter directly affects energy transfer or control effectiveness. For example, the duty cycle of a heat sink can be greater than or equal to 0 and less than or equal to 1.

[0040] In one example, when there are multiple heat source devices, the temperature parameter sequence may include the temperature parameters corresponding to each of the multiple heat source devices at multiple moments in a future time period. Similarly, the preset temperature parameter threshold sequence may include the temperature parameter thresholds corresponding to each of the multiple heat source devices. Furthermore, multiple heat source devices may dissipate heat through a single heat sink. In this case, the control quantity parameter sequence may include the control quantity parameters corresponding to the heat sink at multiple moments in a future time period. Multiple heat source devices may dissipate heat through multiple heat sinks. In this case, the control quantity parameter sequence may include the control quantity parameters corresponding to each of the multiple heat sinks at multiple moments in a future time period. To facilitate description of the technical solutions of this application, the temperature control method will be described below using a single heat sink (the number of heat source devices may be one or more) as an example. The case of multiple heat sinks is similar and will not be further described to avoid repetition.

[0041] In one example, a cost function is constructed based on the difference between the temperature parameter sequence of the heat source device in a future time period and a preset temperature parameter threshold sequence, as well as the control parameter sequence of the heat sink device. The purpose of the cost function is to predict the control parameter in the future time period, striving to minimize the value of the control parameter and the difference between the temperature parameter and the temperature parameter threshold within the future time period. This allows the performance of the heat source device to be fully utilized while preventing the temperature parameter of the heat source device from exceeding the temperature parameter threshold and minimizing the energy consumption of the heat sink device. Here, when the temperature parameter of the heat source device is less than the temperature parameter threshold, the closer the temperature parameter of the heat source device is to the temperature parameter threshold, the greater the degree of utilization of the heat source device's performance. Therefore, based on the cost function, the control parameter sequence can be predicted and optimized to obtain an optimized control parameter sequence. In one example, the optimized control parameter sequence may include optimized control parameters corresponding to multiple moments in the future time period.

[0042] In one example, the cost function can be reasonably constructed based on the above purpose. For example, the model predictive control (MPC) method or other suitable methods can be used to optimize the control parameter sequence of the heat dissipation device in the future time period based on the cost function to obtain the optimized control parameter sequence.

[0043] In one example, based on the current temperature parameter sequence of the heat source device and / or the current control parameter sequence of the heat dissipation device, the control parameter sequence of the heat dissipation device in the future time period can be optimized based on the cost function to obtain an optimized control parameter sequence.

[0044] 220: Acquire the real-time temperature of the heat source device, obtain a real-time temperature parameter sequence, and determine the real-time control quantity parameters of the heat dissipation device according to the real-time temperature parameter sequence and a preset temperature parameter threshold sequence.

[0045] Specifically, the real-time temperature of the heat source device can be obtained through a temperature sensor.

[0046] In one example, the temperature sensor can include a thermistor. In another example, a temperature sensor can be placed in a heat-sensitive area of ​​a heat source device based on finite element / finite volume simulation methods. The temperature difference between the temperature sensor and the heat source device can be less than or equal to 4°C. For heat source devices with high thermal risks, a dual temperature sensor cross-validation method can be used to obtain real-time temperature to improve the reliability of temperature parameters.

[0047] In one example, the number of heat source components may be one or more, and the real-time temperature parameter sequence may include the real-time temperature of each heat source component. Based on the real-time temperature parameter sequence and the preset temperature parameter threshold sequence, the real-time control parameter of the heat dissipation component may be determined. For example, when the real-time temperature of a heat source component exceeds the temperature parameter threshold of the heat source component, the real-time control parameter may be increased; when the real-time temperature is significantly lower than the temperature parameter threshold, the real-time control parameter may be decreased.

[0048] In one example, a proportional-integral (PI) control algorithm, a proportional-integral-derivative (PID) control algorithm, or other suitable control algorithm may be used to determine the real-time control quantity parameter based on the difference between the real-time temperature parameter sequence and the preset temperature parameter threshold sequence.

[0049] 230: Determine target control quantity parameters based on the optimized control quantity parameter sequence and the real-time control quantity parameters.

[0050] In one example, the optimized control quantity parameters and real-time control quantity parameters at each moment in the optimized control quantity parameter sequence can be weightedly calculated to obtain the target control quantity parameters. The weight values ​​of the optimized control quantity parameters and the real-time control quantity parameters can be set according to actual needs.

[0051] In one example, based on the historical temperature changes of the heat source device, such as the temperature gradient at the current moment, one of the first optimized control parameter and the real-time control parameter in the optimized control parameter sequence can be selected as the target control parameter.

[0052] 240: Control the operation of the heat dissipation device based on the target control quantity parameter to achieve temperature control of the heat source device.

[0053] In one example, the principle for selecting the target control quantity parameter may be to prioritize controlling the temperature of the heat source device to be lower than or equal to the temperature parameter threshold to avoid damage to the heat source device, and then control the difference between the temperature of the heat source device and the temperature parameter threshold to be as small as possible and control the control quantity parameter to be as small as possible, that is, to maintain a balance between the difference and the control quantity parameter.

[0054] In one example, if the target control parameter is larger than the control parameter at the previous moment, the temperature of the heat source device may decrease. If the target control parameter is smaller than the control parameter at the previous moment, the temperature of the heat source device may increase to a certain extent.

[0055] Embodiments of the present application provide a temperature control method that, by predicting control parameters of a heat sink, can predict or optimize the control parameters of the heat sink for a future time period in advance. Since the control parameters can affect the temperature parameters of the heat source device, the optimized control parameters can be used to control the temperature of the heat source device to minimize exceeding a temperature parameter threshold during the future time period. This enables proactive control of the heat source device's temperature, enabling early response to temperature rises, reducing the probability of emergency braking and frequent heat sink activation, thereby reducing system energy consumption. Furthermore, by proactively controlling the heat source device's temperature, the temperature parameter threshold can be set more accurately, eliminating the need for empirical estimation. Furthermore, by determining the real-time control parameters of the heat sink based on the real-time temperature of the heat source device, determining target control parameters based on the optimized control parameters and the real-time control parameters for the future time period, and controlling the operation of the heat sink based on the target control parameters, the timely temperature control can be improved based on proactive temperature control of the heat source device, thereby addressing sudden temperature changes.

[0056] According to one embodiment of the present application, step 230 may include: adjusting the weight value of the first optimized control quantity parameter and the weight value of the real-time control quantity parameter in the optimized control quantity parameter sequence according to the temperature gradient corresponding to the heat source device, and performing weighted calculation on the first optimized control quantity parameter and the real-time control quantity parameter to obtain the target control quantity parameter.

[0057] In one example, the temperature gradient can, to a certain extent, represent the rate of change of the heat source device's temperature at the current moment. If the temperature gradient is greater than or equal to the temperature gradient threshold, it indicates that the heat source device's temperature has undergone a sudden change, and timely temperature control is necessary. Therefore, the weight of the real-time control parameter can be increased, while the weight of the optimized control parameter can be decreased. If the temperature gradient is less than the temperature gradient threshold, it indicates that the heat source device's temperature has not undergone a sudden change, and advance control can be primarily employed. Therefore, the weight of the real-time control parameter can be decreased, while the weight of the optimized control parameter can be increased.

[0058] In one example, the optimized control quantity parameter sequence includes optimized control quantity parameters at multiple moments, and the first optimized control quantity parameter in the optimized control quantity parameter sequence can be the optimized control quantity parameter corresponding to the first moment among the multiple moments. As time changes, the control quantity parameter sequence of the heat dissipation device in the future time period can be optimized in real time based on the cost function to obtain the optimized control quantity parameter sequence. That is, an optimized control quantity parameter sequence can be obtained at each optimization moment, and the first optimized control quantity parameter in the optimized control quantity parameter sequence can be selected as the optimized control quantity parameter at the current optimization moment. At the next optimization moment, the first optimized control quantity parameter in the optimized control quantity parameter sequence corresponding to the next optimization moment can be used as the optimized control quantity parameter at the next optimization moment, and so on.

[0059] When determining the target control quantity parameter at the current optimization moment, a weighted calculation may be performed on the real-time control quantity parameter and the first optimized control quantity parameter in the optimized control quantity parameter sequence at the current optimization moment.

[0060] Taking the first optimized control parameter in the optimized control parameter sequence as the optimized control parameter at the current optimization moment can improve the smoothness of the change of the control parameter during the entire temperature control process, thereby avoiding damage to the device life caused by sudden temperature changes of the heat source device.

[0061] In this embodiment, by adjusting the weight value of the first optimized control parameter and the weight value of the real-time control parameter in the optimized control parameter sequence according to the temperature gradient corresponding to the heat source device, and performing weighted calculation on the first optimized control parameter and the real-time control parameter, the target control parameter is obtained. This can improve the timeliness of temperature control on the basis of early control of the temperature of the heat source device, thereby coping with sudden temperature changes.

[0062] According to an embodiment of the present application, according to the temperature gradient corresponding to the heat source device, the weight value of the first optimized control quantity parameter and the weight value of the real-time control quantity parameter in the optimized control quantity parameter sequence are adjusted, and the first optimized control quantity parameter and the real-time control quantity parameter are weightedly calculated to obtain the target control quantity parameter, including: according to the optimized temperature parameter sequence corresponding to the optimized control quantity parameter sequence, the residual between the real-time temperature of the heat source device and the optimized temperature parameter in the optimized temperature parameter sequence, and at least one of the standard deviations of the residuals, and the temperature gradient corresponding to the heat source device, adjusting the weight value of the first optimized control quantity parameter, the weight value of the real-time control quantity parameter and the weight value of the feedforward control quantity parameter, and performing weighted calculation on the first optimized control quantity parameter, the real-time control quantity parameter and the feedforward control quantity parameter to obtain the target control quantity parameter, wherein the feedforward control quantity parameter is determined based on the heat consumption of the heat source device.

[0063] Specifically, the feedforward control parameter may be determined based on the heat consumption of the heat source device. In one example, the heat consumption of the heat source device may represent the heat consumption of the heat source device at the current moment. For example, the value of the feedforward control parameter may be positively correlated with the heat consumption of the heat source device.

[0064] In one example, feedforward control parameters can be derived based on feedforward control. Feedforward control can quickly respond to known disturbances, but it provides a precise model and cannot handle unknown disturbances. The model here refers to the resistor-capacitor (RC) equivalent circuit model used to construct the cost function or predict the temperature of the heat source device. Details are described below. When the model contains errors, steady-state errors cannot be eliminated. The MPC method can be used to determine the optimal control parameter sequence. The MPC method can handle multivariable, constrained, long-term optimization, but it is computationally complex, model-dependent, slow to respond to high-frequency disturbances, and has limited real-time performance. The PI method can be used to determine real-time control parameters. The PI method offers robust control and can eliminate steady-state errors, but it has a delayed response, cannot handle complex motions, and carries the risk of overshoot. Temperature control systems may contain measurable dynamic disturbances, such as sudden load changes and unknown ambient temperature drift. At the same time, the system must meet multi-objective optimization requirements, balancing performance requirements such as temperature accuracy, energy efficiency (e.g., cooling power consumption), and safety (e.g., temperature limits). Furthermore, the model also contains uncertainties, such as aging of the heat transfer medium, aging of the fan, and speed fluctuations. If feedforward control is used alone, model errors will lead to insufficient feedforward compensation, and the temperature will continue to deviate from the set value, such as the temperature parameter threshold, and cannot be corrected. If the MPC method is used alone, under high-frequency disturbances, the MPC rolling optimization delay may cause overshoot, that is, the temperature will temporarily exceed the limit, shortening the equipment life. PI is a feedback control. When faced with a sudden load, PI must wait for the temperature to rise before responding, which can lead to overshoot. For example, if the chip threshold is 105°C, the PI-controlled temperature may first rise to 110°C and then fall back to 105°C, resulting in poor dynamic performance and system instability.

[0065] Therefore, in this embodiment, the control amount parameters obtained by the three control methods can be weightedly calculated to make up for the shortcomings of each control method, thereby improving the reliability of the target control amount parameters.

[0066] In one example, the first optimized control parameter can be expressed as , the real-time control parameter can be expressed as , the feedforward control parameter can be expressed as . and The calculation formula is as follows.

[0067]

[0068]

[0069] in, and are the proportional and integral gains respectively. It is the difference or deviation between the real-time temperature parameter and the temperature parameter threshold. It can be understood as the current moment. It can be understood as a time variable. is the heat consumption of the heat source device. is the feedforward gain, which can be calibrated through experiments. The target control parameter can be expressed as:

[0070]

[0071] , and The sum of the three is 1, and the values ​​of the three can change over time.

[0072] In one example, the three weight values ​​can be determined based on the temperature gradient corresponding to the heat source device, or based on the temperature gradient corresponding to the heat source device and other factors. The other factors can be set according to specific scenarios and actual needs. For example, the other factors may include at least one of an optimized temperature parameter sequence corresponding to the optimized control parameter sequence, a residual between the real-time temperature of the heat source device and the optimized temperature parameter in the optimized temperature parameter sequence, and a standard deviation of the residual.

[0073] In this embodiment, by performing weighted calculation on the control quantity parameters obtained by the three control methods, the rationality and reliability of the target control quantity parameters can be improved, thereby fully utilizing the performance of the heat source device while protecting the life of each device as much as possible.

[0074] According to one embodiment of the present application, when the temperature gradient corresponding to the heat source device is greater than or equal to the temperature gradient threshold, the robot where the heat source device is located is in a transient working condition; when the temperature gradient corresponding to the heat source device is less than the temperature gradient threshold and the residual between the real-time temperature of the heat source device and the optimized temperature parameter is less than or equal to the first residual threshold, the robot is in a steady-state working condition; when the optimized temperature parameter is greater than or equal to the safety threshold, the robot is in a safe working condition; when the standard deviation of the residual is greater than or equal to the standard deviation threshold, the robot is in a parameter mismatch working condition, and the parameter mismatch working condition is used to characterize that the parameters in the cost function do not match the current state of the robot.

[0075] The weight value of the feedforward control parameter under transient conditions is greater than the weight value of the feedforward control parameter under steady-state conditions, safety conditions and parameter mismatch conditions; the weight value of the first optimized control parameter under safety conditions is greater than the weight value of the first optimized control parameter under transient conditions, steady-state conditions and parameter mismatch conditions; the weight value of the real-time control parameter under parameter mismatch conditions is greater than the weight value of the real-time control parameter under transient conditions, steady-state conditions and safety conditions.

[0076] Specifically, the temperature gradient corresponding to the heat source device can be expressed as , the residual between the real-time temperature of the heat source device and the optimized temperature parameter can be expressed as The standard deviation of the residuals can be expressed as The standard deviation threshold can be expressed as 3 , It can represent the historical residual standard deviation.

[0077] In one example, for different scenarios, the robot's working conditions can be divided into transient working conditions, steady-state working conditions, safe working conditions, and parameter mismatch working conditions.

[0078] In one example, the quantitative standards for the above four working conditions are as follows:

[0079] Transient conditions: Temperature gradient threshold, for example, the temperature gradient threshold is 2°C / s. , and They are 0.2, 0.1, and 0.7 respectively, that is, at this time, the feedforward control dominates the temperature control process and can suppress rapid disturbances.

[0080] Steady-state conditions: Temperature gradient threshold and The first residual threshold, such as the first residual threshold temperature gradient threshold is 1°C, , and They are 0.6, 0.2, and 0.2 respectively. That is, long-term optimization is performed based on MPC and fine-tuning is performed based on PI.

[0081] Safe working conditions: Optimized temperature parameters Safety threshold, such as the safety threshold is 90℃, , and They are 0.8, 0.1, and 0.1 respectively, that is, forced cooling is performed based on MPC, and assisted by PI and feedforward control.

[0082] Parameter mismatch conditions: 3 , , and They are 0.3, 0.6, and 0.1 respectively, which means that the correction is mainly based on PI at this time.

[0083] In one example, a dynamic correction factor may be further set, and the weight value may be fine-tuned based on real-time indicators, such as temperature gradient and the size of the standard deviation of the residual. , and The calculation formula is as follows:

[0084]

[0085]

[0086]

[0087] in, , as well as These three correction factors can be determined according to the following formula.

[0088]

[0089]

[0090]

[0091] in, , as well as is the corresponding gain coefficient, which can be obtained through experimental calibration. Specifically, the larger the residual, the lower the MPC weight, that is, the lower the model credibility; the larger the residual, the stronger the PI correction; the faster the temperature changes, the greater the weight of the feedforward control.

[0092] According to one embodiment of the present application, the optimized control quantity parameter sequence corresponds to an optimized temperature parameter sequence, the optimized temperature parameter sequence includes optimized temperature parameters of the heat source device at multiple moments in a future time period, the heat transfer process of the heat source device corresponds to a thermal resistance-heat capacitance equivalent circuit, and the temperature parameter sequence is determined based on the thermal resistance values ​​of multiple thermal resistors in the thermal resistance-heat capacitance equivalent circuit. The temperature control method further includes: calculating the real-time temperature gradient of the heat source device at a historical moment, and the optimized temperature gradient corresponding to the optimized temperature parameters of the heat source device at the historical moment; obtaining a temperature gradient residual based on the real-time temperature gradient and the optimized temperature gradient; and updating at least some of the thermal resistance values ​​in the cost function when the temperature gradient residual is greater than or equal to a second residual threshold.

[0093] Specifically, the heat transfer process of the heat source device corresponds to a thermal resistance-heat capacitance equivalent circuit. The thermal resistance-heat capacitance equivalent circuit can be called a thermal resistance-heat capacitance equivalent circuit model, which can be abbreviated as an RC model. The model can also be called a temperature prediction model.

[0094] In one example, a thermal resistance-heat capacitance equivalent circuit may include multiple thermal resistances.

[0095] In one example, for a single heat source device, the temperature rise curve of the single heat source device can be tested separately under different heat source heating (different operating and control conditions) and different heat dissipation boundary conditions (corresponding to different fan speeds). For example, the temperature rise curve can be collected through a temperature sensor, and the heat consumption curve of the single heat source device can be collected at the same time. For example, if the heat source device is a CPU, its voltage curve or current curve can be collected as the heat consumption curve. For example, the horizontal axis of the temperature rise curve (Tt) can be time t, and the vertical axis can be temperature T; the horizontal axis of the heat consumption curve (Qt) can be time t, and the vertical axis can be heat consumption Q. Based on the temperature rise curve (Tt) and the heat consumption curve (Qt), a fitting curve (TQ) can be obtained. Based on the fitting curve (TQ), the thermal resistance R and heat capacitance C in the thermal resistance-heat capacitance equivalent circuit can be obtained.

[0096] In one example, for a multi-heat source device, such as a joint motor whose primary heat sources include coils and MOSFETs (hereinafter referred to as MOS), it is necessary to control the heat generation of a single heat source device. The coil's temperature sensor collects the coil's temperature curve, and the MOS's temperature sensor collects the MOS's temperature curve. Simultaneously, the corresponding heat dissipation curves for the coil and MOS are collected. A fitting curve (TQ) is generated based on the temperature rise curve (Tt) and the heat dissipation curve (Qt). The fitting curve (TQ) can be used to determine the thermal resistance R and thermal capacitance C in the thermal resistance-heat capacitance equivalent circuit.

[0097] For the collected data of the heat source device, the corresponding RC network order can be selected, and the model order can be selected according to the actual heat dissipation path. For example, for the heat dissipation path: heat source device-heat conducting medium-housing-environment, a 3rd order RC network can be selected to fit the residual R 2 ≥0.99 is used as the standard to reduce the complexity of model parameters. For multiple heat source devices, the superposition principle can be used to obtain the RC network model of the temperature sensor. When there are two heat source devices, the coil and the MOS, taking the temperature collected by the coil temperature sensor as an example, the calculation formula for the coil temperature parameter (such as the predicted temperature parameter or the optimized temperature parameter) is as follows:

[0098]

[0099] in, Can represent the temperature parameters of the coil, It can represent MOS Joule heat or MOS heat dissipation. It can represent the RC network of MOS to coil, It can represent the Joule heating of the coil or the heat loss of the coil. Can represent ambient temperature parameters.

[0100] It can represent the RC network of the coil itself, or the fan inlet temperature or the ambient temperature.

[0101] In one example, the first term in the above formula represents the coil temperature rise caused by the interaction between the coil and the MOS, that is, the effect of the MOS on the coil temperature rise. The order of the first term is second order, and the heat dissipation path is: MOS - motor housing - coil. i represents the order and can be 1 or 2. It can represent the thermal resistance of MOS-motor housing, It can represent the thermal resistance of the motor housing and coil. Can represent the heat capacity of MOS-motor housing, It can represent the heat capacity of the motor housing-coil.

[0102] The second term represents the temperature rise caused by the coil's own heating. The second term is third-order, with the heat dissipation path being: coil - thermally conductive medium (which can be a thermally graded material, such as heat sink material) - motor housing - ambient. j represents the order, with values ​​of 1, 2, or 3. It can represent the thermal resistance of coil-heat conducting medium. It can represent the thermal resistance of heat conducting medium-motor housing. It can represent the thermal resistance between the motor housing and the environment. It can represent the heat capacity of coil-heat conducting medium. It can represent the heat capacity of the heat conducting medium - motor housing, It can represent the heat capacity of the motor housing-environment.

[0103] above It can represent the temperature parameters of the coil, such as the change in temperature. It can also be used to represent the temperature of the coil. In this case, a third term can be added to the above formula, which can represent the temperature of the coil at the current moment.

[0104] In one example, for a mobile device including only a single heat source device such as a coil, the coupling term can be removed, that is, the first term in the above formula can be removed.

[0105] In one example, each temperature parameter in the temperature parameter sequence may be determined based on a thermal resistance value of a related thermal resistor in a thermal resistance-heat capacitance equivalent circuit.

[0106] Similarly, the calculation formula for the temperature parameters of the MOS, such as the optimized temperature parameter sequence, is similar to the above formula and will not be described again here to avoid repetition.

[0107] It should be understood that the above formula can be selected based on the structure and hardware layout of the entire computing device. The heat capacity in the above formula is related to the physical properties of the structure. The above formula can be used to predict the temperature parameters of the heat source device at a future time.

[0108] In one example, the actual heat dissipation path between the MOS and the coil can be: MOS - thermal conductive medium - motor housing - thermal conductive medium - coil, which is a fourth-order RC model. To simplify the model and reduce the computational complexity, it can be simplified to a second-order RC model: MOS - motor housing - coil.

[0109] In one example, a thermal dissipation curve (based on a thermal dissipation model, which, for a motor, includes Joule heating of the windings and heat dissipation of the driver MOS) can be input into an RC model based on a sliding time window, for example, a 30-second future time period. This can yield a predicted temperature parameter sequence, or an optimized temperature parameter sequence. In one example, the temperature parameter sequence can be updated in real time.

[0110]

[0111] It can represent the optimized temperature parameter sequence, It can be a matrix, for example, a 3*3 matrix, where three rows represent three heat source devices, and each row represents the temperature parameters of each heat source device at multiple moments. yes The transpose of . The calculation formula and the above The calculation formula is essentially the same, except that Including the optimized temperature parameters of a heat source device at multiple times in the future time period, It may include optimized temperature parameters of multiple heat source devices at multiple moments in a future time period. It can represent the heat consumption of multiple heat source devices at multiple moments in the future time period. In one example, Can be predicted by motion control, So we can predict the temperature, that is, For example, motion control can affect the value of the control parameter, and the value of the control parameter can in turn affect the value of the temperature parameter.

[0112] In one example, the cost function can be expressed as follows:

[0113]

[0114] in, It can represent the control parameter sequence, It can represent a sequence of preset temperature parameter thresholds, is the weight value of the control parameter, which is used to prevent over-adjustment. In one example, the constraints of the cost function include: each control parameter in the control parameter sequence satisfy In one example, the optimization problem can be solved once in each control cycle, and only the first step control quantity parameter in the optimization control quantity parameter sequence is applied. , that is, the first optimized control parameter mentioned above In one example, the cost function can minimize the difference between the optimized temperature parameter sequence (or predicted temperature parameter sequence) and the preset temperature parameter threshold sequence, while minimizing the optimized control parameter sequence, that is, maintaining a balance between the difference and the optimized control parameter sequence.

[0115] Taking a single heat source device as an example, the optimized temperature parameter (or predicted temperature parameter) for this heat source device is 100°C, and the temperature parameter threshold for this heat source device is 120°C. A primary goal of optimizing the cost function is to keep the temperature parameter of the heat source device as close to the temperature parameter threshold as possible in the future. Generally, if the control parameter value is too large, that is, the motor speed is too high, it results in wasted performance. For example, if the temperature parameter threshold is 120°C, and the temperature parameter of the heat source device is currently controlled at 100°C, there is a 20-degree margin between 100°C and 120°C, which is a waste of performance. Therefore, the speed can be controlled to maintain near 120°C, or close to 120°C, at a more reasonable cost.

[0116] If the speed is increased, the temperature of the heat source components will be reduced to prevent damage to the heat source components, which will increase operating costs and shorten the life of the fan. If the speed is too low, the temperature of the heat source components will exceed the temperature parameter threshold, which is also not good.

[0117] The best way is to control the temperature of the heat source device near the temperature parameter threshold.

[0118] In one example, the motor's current and / or voltage determine heat dissipation, which is controlled by the robot's motion control. The robot's movements may differ at different times, such as walking at one moment and dancing at the next. Consequently, the robot's power consumption or heat dissipation, and therefore its temperature, may differ from the first moment to the second. In this case, real-time control parameters must be provided to the fan to stabilize the robot's temperature and prevent significant fluctuations. Each temperature parameter threshold in the preset temperature parameter threshold sequence can be associated with the characteristics of the corresponding heat source component.

[0119] By optimizing temperature to control power consumption or heat output, this is a process of dynamic whole-machine energy planning. Through this process, the temperature of the heat source device can be maintained near the critical value (temperature parameter threshold), thereby passively maintaining the temperature of the heat source device.

[0120] In one example, taking a single heat source device as an example, the first term in the above formula is the minimum residual sum of squares between the optimized temperature parameter of the heat source device and the temperature parameter threshold, which is used to ensure that the temperature of the heat source device is as close to the target value (temperature parameter threshold) as possible. The second term in the above formula is the suppression control parameter ( ) increase, avoiding frequent and large-scale operation of heat dissipation components such as fans, thereby reducing energy consumption and mechanical wear.

[0121] In the above formula, N is the prediction time domain, and the prediction period is N*dt. Here, the prediction period can represent the aforementioned future time period. Alternatively, N can represent the time step. For example, if the prediction period is 30 seconds and three samples are taken within 30 seconds, then the value of N can be 3.

[0122] In one example, N can be selected based on the system thermal inertia, usually 1 / 3-1 / 2 of the time constant (usually RC). M is the control time domain, usually M=N / 2, to balance the control effect and the amount of calculation. M≤N, the larger N is, the stronger the long-term prediction ability is, but the amount of calculation is increased. The larger M is, the more flexible the control is, but the optimization variables ( ) increases. λ is the regularization weight coefficient, which is used to balance the weight of the temperature tracking error and the size of the control parameter. The larger λ is, the smoother the control is, but the temperature tracking speed may be sacrificed and the response speed may be slower. λ can be obtained through step response testing. The value that makes the overshoot ≤5% and the adjustment time the shortest can be selected. The weight can be adjusted according to the temperature deviation. Adaptive adjustments, such as reducing λ at high temperatures to prioritize temperature safety.

[0123] In one example, the cost function may be constructed based on the MPC method, that is, the process of solving the optimal control quantity parameters may be performed based on the MPC method.

[0124] In one example, the ultimate goal of the cost function is to minimize the value of the first term. The control parameters (such as duty cycle) adopted at each time step may be different. The second term in the cost function is to prevent overshoot by controlling the fan control parameters to minimize the difference between the optimized temperature parameter and the temperature parameter threshold. The optimal control parameter sequence It can be the one that minimizes the cost function .

[0125] The above cost function minimizes the difference between the optimized temperature parameter and the temperature parameter threshold, while also constraining the size of the control parameter. This ensures that the temperature of the heat source component is as close as possible to the target value (temperature parameter threshold) while preventing the control parameter from being too large. This reduces costs, fully utilizes the performance of the heat-generating component, and, to a certain extent, extends the service life of the heat sink and heat source components.

[0126] In one example, control instructions for a mobile device, such as a robot, can determine the voltage and current of the motors of the associated joints. Current determines heat dissipation. Controlled variables can also determine the thermal resistance between the motor housing and the environment, and therefore, the controlled variables can determine the temperature parameters of the heat source device.

[0127] Furthermore, in one example, the thermal resistance and thermal capacity in the above RC model will change as the usage time of the heat source device increases. In order to obtain such changes in a timely manner, the real-time temperature gradient of the heat source device at the historical moment and the optimized temperature gradient corresponding to the optimized temperature parameters of the heat source device at the historical moment can be calculated in real time.

[0128] In one example, the real-time temperature gradient can be a temperature gradient calculated based on the actual temperature at a historical moment. For example, the optimized temperature gradient can be obtained by derivation or other methods. Specifically, real-time temperature parameters at multiple moments can be obtained, and based on the real-time temperature parameters at multiple moments, the real-time temperature parameter at a specific historical moment can be derived to obtain the real-time temperature gradient at the specific historical moment.

[0129] Similarly, the optimized temperature parameters of the heat source device at a historical moment may include temperature parameters obtained at that historical moment based on cost function optimization. Based on the optimized temperature parameters at that historical moment, a corresponding optimized temperature gradient may be obtained. For example, the optimized temperature gradient may be obtained through derivation or other methods. Specifically, the optimized temperature parameters (or predicted temperature parameters) at multiple moments may be obtained. Based on the optimized temperature parameters at multiple moments, the optimized temperature parameter at a specific historical moment may be derived to obtain the optimized temperature gradient at that specific historical moment.

[0130] In one example, the difference between the real-time temperature gradient and the optimized temperature gradient may be used as the temperature gradient residual.

[0131] Optionally, the absolute value of the difference between the real-time temperature gradient and the optimized temperature gradient can be used as the temperature gradient residual. When the temperature gradient residual is greater than or equal to the second residual threshold, it means that the difference between the temperature predicted by the cost function and the actual temperature of the heat source device is large, that is, the predicted temperature accuracy is low. The reason for this situation may be that the thermal resistance value of at least part of the thermal resistors in the above-mentioned RC model has changed. In order to ensure the accuracy of the subsequent predicted temperature, it is necessary to update or correct the thermal resistance value of at least part of the thermal resistors in the RC model, that is, to update at least part of the thermal resistance value in the cost function (or the calculation formula of the temperature parameters of the heat source device).

[0132] In one example, the thermal resistance value that has a greater impact on the predicted temperature may be updated, thereby saving calculation effort.

[0133] For example, using the coil and MOS as two heat source devices, the thermal resistances that have a significant impact on the predicted temperature (i.e., the value of the optimized temperature parameter) may include: the interactive thermal network thermal resistance and the thermal resistance from the motor housing to the environment. Specifically, assuming the heat dissipation path of the coil is: coil-motor housing-environment, and the heat dissipation path of the MOS is: MOS-motor housing-environment. The interactive thermal network thermal resistance may include the thermal resistance of the coil-motor housing (which is equal to the thermal resistance of the motor housing-coil). , thermal resistance of motor housing-MOS (equal to thermal resistance of MOS-motor housing) The thermal resistance from the motor housing to the environment may include the thermal resistance from the motor housing to the environment in the "coil-motor housing-environment" heat dissipation path. The thermal resistance between the motor housing and the environment in the "MOS-motor housing-environment" heat dissipation path .

[0134] In one example, the size of the heat capacitance has little effect on the value of the optimized temperature parameter, so the heat capacitance value in the above cost function (or the calculation formula of the temperature parameter of the heat source device) does not need to be updated, which can save computing resources and improve update efficiency.

[0135] In one example, the real-time temperature gradient can be expressed as , the optimized temperature gradient can be expressed as , the temperature gradient residual can be expressed as , .

[0136] In one example, the second residual threshold can be set according to actual needs. When the temperature gradient residual is greater than or equal to the second residual threshold, the RC model can be triggered to adaptively adjust the thermal resistance value of at least part of the thermal resistor. In one example, the second residual threshold can be σ can be the standard deviation of the temperature gradient residual based on historical data.

[0137] Furthermore, when the update is triggered continuously, for example, the number of times the RC model updates the thermal resistance value within a preset time period is greater than or equal to a preset threshold, the threshold can be tightened according to the following rules, such as increasing the update frequency, that is, reducing the second residual threshold:

[0138] ,

[0139] Where t is time, such as the continuous operating time of the entire machine, device, or component, n can be the nth time point, and n-1 can be the n-1th time point. For example, at the 8000th time point, the entire machine has been running for 2 hours, so t is 2 hours. The second residual threshold at the 8000th time point can be determined based on the second residual threshold at the 7999th time point. In one example, the time point can be determined based on the sampling frequency. For example, if three points are collected every 10 minutes, the 20th minute is the sixth time point.

[0140] The half-life is . Over time, the performance of the heat source device can decline exponentially, meaning that the performance decays at an increasingly rapid rate. Accordingly, the second residual threshold should be tightened and updated more frequently. That is, the second residual threshold at the nth time point is smaller than the second residual threshold at the n-1th time point. Tightening the threshold here is to match the performance decay pattern of the heat source device.

[0141] In one example, The larger the value, the smaller the threshold tightening and the lower the update frequency. The smaller the value, the greater the threshold tightening and the higher the update frequency. In order to balance computing resources and model accuracy, you can choose a suitable ,For example, The possible value is 30min.

[0142] In this example, increasing the update frequency can ensure the accuracy of the model. In addition, by selecting an appropriate update frequency, computing resources can be saved while ensuring the accuracy of the model.

[0143] In this embodiment, when the temperature gradient residual is greater than or equal to the second residual threshold, at least part of the thermal resistance values ​​in the cost function are updated. This can improve the accuracy of the temperature parameters predicted by the RC model and ensure the rationality of the optimized control quantity parameters obtained based on the cost function.

[0144] According to one embodiment of the present application, at least part of the thermal resistance values ​​in the cost function are updated, including: obtaining state parameters, wherein the state parameters include temperature parameters of the heat source device, ambient temperature parameters, and at least part of the thermal resistance values; using the Kalman filtering method to update part of the thermal resistance values ​​based on the state parameters.

[0145] In one example, the state parameter can be expressed in the form of a vector, which can also be called a state vector. The state vector X is expressed as follows:

[0146]

[0147] in, It can represent the temperature of the temperature sensor corresponding to the coil, that is, the temperature parameter of the coil. It can represent the temperature of the temperature sensor corresponding to the MOS, that is, the temperature parameter of the MOS. Can represent ambient temperature parameters.

[0148] , , as well as The specific meaning of can be found in the description of the above embodiment. The thermal resistance values ​​of these four thermal resistors are strongly correlated with convective heat transfer, and the convective thermal resistance in the system has a significant impact on the optimization results. Therefore, to reduce the amount of calculation, these four thermal resistance parameters can be used as the state vector for update. The state vector contains not only temperature dynamics but also thermal network parameters, representing a joint estimate of parameters and state. In other examples, RC parameters of second order and above can be ignored to reduce the amount of calculation.

[0149] The following describes the specific process of updating some thermal resistance values ​​based on state parameters using the Kalman filter method.

[0150] In one example, the nonlinear state equation in the Kalman filter method is expressed as follows:

[0151]

[0152] Wherein, U represents the heat consumption of the heat source device, and specifically can represent the heat consumption of each heat source device. is a nonlinear state transfer equation, for example, it can be used to characterize X under different U. w(t) is the process noise (usually assumed to be Gaussian distributed, its covariance matrix is ​​Q, and the initial value can be obtained through experimental calibration or model derivation), which is used to describe the temperature (such as ) and parameters (such as ) is a dynamic process that evolves over time, that is, the derivative of the state vector with respect to time ( For the coil, the time rate of change of the coil temperature can be obtained by taking the derivative of the coil RC network equation with respect to time. For unobservable state variables, such as , , ,or (Assume it is a constant) is a random walk model, that is, d / dt=0.

[0153] In one example, the observation vector in the Kalman filter method is as follows:

[0154]

[0155] The observation matrix is ​​H, which is used to extract the observable node temperature:

[0156]

[0157] The observation equation is:

[0158]

[0159] v(t) is the observation noise, which is usually assumed to be Gaussian distributed, and its covariance matrix is ​​R.

[0160] Calculate the residual between the observed temperature (real-time temperature parameter) and the predicted temperature (optimized temperature parameter) :

[0161]

[0162] It can represent the observed temperature of MOS, It can represent the observed temperature of the coil, It can represent the predicted temperature of MOS, It can represent the predicted temperature of the coil.

[0163] Distribute the covariance matrix R according to the accuracy of the temperature sensor:

[0164]

[0165] in, For the accuracy of the MOS temperature sensor, is the accuracy of the coil temperature sensor.

[0166] Furthermore, the state transition Jacobian matrix F can be calculated:

[0167]

[0168] The Jacobian matrix of the observation equation is calculated as follows:

[0169]

[0170] That is, the derivative of the i-th element in the observation matrix with respect to the j-th state variable is obtained.

[0171] In one example, the Extended Kalman Filter (EKF) iterative update process is as follows:

[0172] In the prediction phase, the prediction status is represented as:

[0173] The forecast covariance is expressed as:

[0174] Xk|k-1 is the predicted state at time k based on the state k-1.

[0175] Fk-1 is the transfer Jacobian matrix of F at time k-1.

[0176] Pk|k-1 is the prediction covariance matrix, which represents the state estimation uncertainty, that is, covariance propagation.

[0177] Use the model to predict the current state and quantify the uncertainty of the prediction. Calculate the Jacobian matrix: including the state transition Jacobian matrix and the observation Jacobian matrix.

[0178] Linearize nonlinear models for covariance propagation and Kalman gain calculation.

[0179] Update step, Kalman gain calculation and state correction:

[0180] Kalman gain

[0181] Status Update

[0182] Covariance Update

[0183] Among them, K k is the Kalman gain, which measures the confidence of prediction and measurement, and I is the identity matrix. The initial value of the covariance matrix P can be obtained by model derivation or based on the initial state. The predicted state is corrected by sensor data to reduce the uncertainty of model estimation, that is, the unmeasurable parameter R in the prediction model is updated, where Contains the updated state parameter R.

[0184] In other examples, other suitable methods may be used to update at least part of the thermal resistance value.

[0185] In this embodiment, the Kalman filter method is used to update some thermal resistance values ​​based on the state parameters, which can improve the accuracy of the thermal resistance value update and further improve the optimization effect of the optimized control quantity parameters.

[0186] According to one embodiment of the present application, the temperature control method also includes: when the gap between the preset temperature parameter threshold sequence and the optimized temperature parameter sequence corresponding to the optimized control quantity parameter sequence meets the preset conditions, the motion instructions of the robot where the heat source device is located are adjusted, and the load of the joint corresponding to the heat source device in the adjusted motion instructions is less than the load of the joint corresponding to the heat source device in the motion instructions before adjustment.

[0187] In one example, the preset condition can be used to indicate that the difference between the temperature parameter threshold of the heat source device in the preset temperature parameter threshold sequence and the corresponding optimized temperature parameter of the heat source device in the optimized temperature parameter sequence is sufficiently small. For example, the preset condition can include a first critical value. If the difference is less than or equal to the first critical value, it indicates that the temperature of the heat source device is at risk of exceeding the temperature parameter threshold at a future time. Therefore, the temperature of the heat source device can be proactively controlled in advance, such as by reducing the load of the heat source device, to prevent overheating and damage to the heat source device.

[0188] Furthermore, if the difference is less than or equal to the first critical value, while reducing the load on the joint corresponding to the heat source device, the operation of the heat dissipation device can also be controlled by optimizing the control parameter. This passive cooling method can maintain the temperature of the heat source device near and below the temperature parameter threshold. In this way, the power management strategy (corresponding to load reduction) and the thermal management strategy (corresponding to optimizing the control parameter) can be combined to ensure the safe operation of the entire machine, save power energy, and extend operating time.

[0189] In one example, the first critical value can be set according to actual needs.

[0190] In one example, the difference between the temperature parameter threshold and the optimized temperature parameter can be used to predict the time it takes for a heat source component to reach the temperature parameter threshold (e.g., overtemperature time). This time can then be used to determine the corresponding power management strategy. For example, if a heat source component takes a long time to reach the temperature parameter threshold, the load on the joint corresponding to the heat source component can be reduced slowly. Otherwise, the load on the joint corresponding to the heat source component can be reduced more quickly.

[0191] In one example, if the optimized temperature parameter is less than a safety threshold, full performance mode can be enabled, allowing the joint corresponding to the heat source device to move at maximum load. If the optimized temperature parameter is greater than or equal to the safety threshold and less than the warning temperature, a reduced power operation mode can be enabled, limiting the peak torque of the joint corresponding to the heat source device. If the optimized temperature parameter is greater than or equal to the warning temperature, an emergency cooling mode can be enabled, suspending high-load movements and activating a strong cooling mode.

[0192] In one example, the robot's task priority and thermal status can be integrated to reasonably distribute the power or load of the joints. Specifically, for a specific joint, the real-time temperature of the corresponding heat source device can be used to calculate the power or load of the joints. , safe temperature , warning temperature As well as the maximum power of the joint, the power of the joint can be reasonably distributed. For example, the power of the joint can be determined by the following formula:

[0193]

[0194] in, is the power of the joint after power distribution, is the maximum power of the joint, is the mission urgency coefficient, which can be obtained through calibration.

[0195] In this embodiment, when the gap between the temperature parameter threshold and the optimized temperature parameter meets the preset conditions, the motion instructions of the robot where the heat source device is located are adjusted so that the load of the joint corresponding to the heat source device in the adjusted motion instructions is less than the load of the joint corresponding to the heat source device in the motion instructions before adjustment. In this way, the load of the joint corresponding to the heat source device can be adjusted in time and automatically by predicting the temperature, thereby improving the stability and safety of the robot's working process.

[0196] According to one embodiment of the present application, the robot includes multiple joints, and the joint corresponding to the heat source device includes the first joint among the multiple joints, wherein the motion instructions of the robot where the heat source device is located are adjusted, including: reducing the load of the first joint and increasing the load of the second joint among the multiple joints.

[0197] In one example, the heat source device may be a device such as a coil or MOS for driving or promoting joint movement.

[0198] In one example, the robot can perform the target action by collaborating with multiple joints, and the multiple joints may include a first joint and a second joint. The heat source device can be used to drive or drive the first joint to move. When the robot controls the movement of multiple joints based on a pre-set motion instruction, if the gap between the temperature parameter threshold corresponding to the heat source device of the first joint and the optimized temperature parameter of the heat source device meets the preset conditions, the load of the first joint needs to be reduced. At this time, in order to ensure the completion of the target action, the load of the second joint can be appropriately increased. It should be understood that in the process of the robot performing the target action, the second joint can be a joint that can withstand a larger load. Therefore, the load of the second joint can be appropriately increased to complete the target action. Assuming that there is a third joint among the multiple joints, the load that the third joint can withstand is smaller in the process of the robot performing the target action. At this time, even if the load of the first joint needs to be reduced, the load of the third joint will not be increased. You can choose to increase the load of the second joint, or choose to delay the completion of the target action.

[0199] In one example, the load of the joint may represent at least one of the velocity, acceleration, bearing force, torque, power, etc. of the joint.

[0200] In one example, a quantitative relationship between joint motion (torque, speed) and motor losses can be established, converting motion commands into heat source inputs to drive a temperature prediction model, i.e., the aforementioned RC model, for temperature prediction. Motion commands can be replanned based on the predicted temperature. For example, based on the predicted upper temperature limit, the maximum allowable joint torque / speed can be inferred to generate safe motion commands (based on which a safe motion trajectory can be derived), thus reducing the need for frequent fan adjustments.

[0201] In this embodiment, when the gap between the temperature parameter threshold corresponding to the heat source device of the first joint and the optimized temperature parameter of the heat source device meets the preset conditions, by reducing the load of the first joint and increasing the load of the second joint, the target action can be completed in a timely manner while protecting the heat source device.

[0202] According to one embodiment of the present application, the load of the first joint is negatively correlated with the temperature difference variable, wherein the temperature difference variable is used to characterize the ratio between the difference between the real-time temperature of the heat source device and the safety temperature corresponding to the heat source device and the difference between the warning temperature corresponding to the heat source device and the safety temperature corresponding to the heat source device.

[0203] In one example, the temperature difference variable can be expressed as follows:

[0204]

[0205] Indicates the real-time temperature of the heat source device. Indicates the safety temperature corresponding to the heat source device. Indicates the warning temperature corresponding to the heat source component.

[0206] In one example, during the thermal management process, multiple joints can be controlled to collaboratively dissipate heat so that the load on each joint is balanced. For example, when the temperature of the heat source device of the first joint approaches the temperature parameter threshold, at least part of the load of the first joint can be transferred to another joint, such as the second joint. This can prevent the first joint from malfunctioning due to overheating while ensuring the timely completion of the target action. In another example, a robot can perform dance movements by controlling multiple joints in the legs. When the heat source device of the hip joint of the leg overheats or its temperature approaches the temperature parameter threshold, part of the load of the hip joint can be distributed to the knee joint.

[0207] In one example, the load factor of the first joint can be expressed as follows:

[0208]

[0209] in, is the load transfer coefficient, which can be adjusted through the operation control algorithm. Can be the current load factor of the first joint, The product of the current load coefficient of the first joint and the maximum load of the first joint is the current load of the first joint.

[0210] In this embodiment, the load of the first joint is associated with the temperature difference variable. When the current temperature of the heat source device of the first joint is too high, the load of the first joint can be adjusted, and the load of the first joint can be quantitatively adjusted. In addition, the load subtracted from the first joint can be transferred to other joints. In this way, the failure of the first joint due to overheating can be avoided, and the timely completion of the target action can be ensured. In actual application scenarios, the solution of this embodiment can be used to reasonably adjust the robot's gait posture in exchange for stronger heat dissipation performance.

[0211] Figure 4 Shown is a flow chart of a temperature control method provided by another exemplary embodiment of the present application. Figure 4 The embodiment is Figure 2 For the example of embodiment, in order to avoid repetition, the same points can be referred to the description in the above embodiment, which will not be repeated here. Figure 4 As shown, the temperature control method may include the following contents.

[0212] 310: Optimizing a control parameter sequence of the heat dissipation device in a future time period based on the cost function to obtain an optimized control parameter sequence.

[0213] Specifically, the specific contents of the cost function, heat dissipation device, future time period, control quantity parameter sequence and optimized control quantity parameter sequence can be found in the description of the above embodiments, and will not be repeated here to avoid repetition.

[0214] 320: Acquire the real-time temperature of the heat source device, obtain a real-time temperature parameter sequence, and determine the real-time control quantity parameters of the heat dissipation device according to the real-time temperature parameter sequence and a preset temperature parameter threshold sequence.

[0215] Specifically, the specific contents of the heat source device, real-time temperature, real-time temperature parameter sequence, preset temperature parameter threshold sequence and real-time control quantity parameters can be found in the description of the above embodiments. To avoid repetition, they will not be repeated here.

[0216] 330: According to the temperature gradient corresponding to the heat source device, adjust the weight value of the first optimized control parameter and the weight value of the real-time control parameter in the optimized control parameter sequence, and perform weighted calculation on the first optimized control parameter and the real-time control parameter to obtain the target control parameter.

[0217] Specifically, the specific contents of the temperature gradient, the first optimized control parameter and the target control parameter can be found in the description of the above embodiment, and will not be described again here to avoid repetition.

[0218] 340: Control the operation of the heat dissipation device based on the target control quantity parameter to achieve temperature control of the heat source device.

[0219] 350: Calculate the real-time temperature gradient of the heat source device at the historical moment, and the optimized temperature gradient corresponding to the optimized temperature parameter of the heat source device at the historical moment.

[0220] Specifically, the specific contents of the historical moments, the real-time temperature gradient, and the optimized temperature gradient can be found in the description of the above embodiments, and will not be described again here to avoid repetition.

[0221] 360: Get the temperature gradient residual based on the real-time temperature gradient and the optimized temperature gradient.

[0222] Specifically, the specific content of the temperature gradient residual can be found in the description of the above embodiment, and will not be described again here to avoid repetition.

[0223] 370 : When the temperature gradient residual is greater than or equal to a second residual threshold, update at least part of the thermal resistance values ​​in the cost function.

[0224] Specifically, the specific contents of the second residual threshold and at least part of the thermal resistance value can be found in the description of the above embodiment, and will not be described again here to avoid repetition.

[0225] 380: When the gap between the preset temperature parameter threshold sequence and the optimized temperature parameter sequence corresponding to the optimized control parameter sequence meets the preset conditions, the motion instruction of the robot where the heat source device is located is adjusted, and the load of the joint corresponding to the heat source device in the adjusted motion instruction is less than the load of the joint corresponding to the heat source device in the motion instruction before adjustment.

[0226] Specifically, the specific contents of optimizing the temperature parameter sequence, preset conditions and joint load can be found in the description of the above embodiments, and will not be repeated here to avoid repetition.

[0227] It should be understood that the execution order of the above steps can be adjusted according to actual needs.

[0228] The temperature control method provided in the embodiment of the present application can realize the temperature rise prediction of the heat source device by the RC model. The order of the model corresponds to the physical model, which is simple and reliable, and improves the reliability of the temperature control process. Specifically, based on the RC model, high-precision prediction of the temperature of the heat source device can be quickly achieved, and the parameters of the RC model correspond to the real model and are interpretable. Furthermore, the temperature control method provided in the embodiment of the present application can integrate thermal management with the whole machine operation and control logic, and integrate MPC with PI algorithms to achieve a temperature control process without lag and overshoot. Moreover, the embodiment of the present application can use the extended Kalman filter algorithm to realize adaptive correction of the model, and form a closed loop with the operation and control to improve the compatibility of the temperature control method. In addition, the temperature control method provided in the embodiment of the present application has strong generalization and can perform self-model adjustment according to the differences between different robots.

[0229] Exemplary devices

[0230] The embodiment of the present application also provides a robot, which includes a control module, which is used to execute the above Figure 2 or Figure 3 The temperature control method provided in the embodiment.

[0231] In some embodiments, the robot can be a wheeled robot, a humanoid robot, a tracked robot, or other types of robots. For example, the robot can be a robot comprising an arm, such as a robot comprising one arm or two arms.

[0232] Figure 4 The figure shows a schematic diagram of the structure of a temperature control device provided by an exemplary embodiment of the present application. Figure 4 As shown, the temperature control device 400 includes: an optimization module 410 , an acquisition module 420 , a determination module 430 and a control module 440 .

[0233] The optimization module 410 is configured to optimize the control parameter sequence of the heat sink in a future time period based on a cost function, thereby obtaining an optimized control parameter sequence. The acquisition module 420 is configured to obtain the real-time temperature of the heat source component, thereby obtaining a real-time temperature parameter sequence. The real-time control parameter sequence and a preset temperature parameter threshold sequence are then used to determine the real-time control parameter of the heat sink. The determination module 430 is configured to determine the target control parameter based on the optimized control parameter sequence and the real-time control parameter. The control module 440 is configured to control the operation of the heat sink based on the target control parameter to achieve temperature control of the heat source component.

[0234] Embodiments of the present application provide a temperature control device that, by predicting control parameters of a heat sink, can predict or optimize the control parameters of the heat sink for a future time period in advance. Since the control parameters can affect the temperature parameters of the heat source device, the optimized control parameters can be used to control the temperature of the heat source device to minimize exceeding a temperature parameter threshold during the future time period. This enables proactive control of the heat source device's temperature, enabling early response to temperature rises, reducing the probability of emergency braking and frequent heat sink activation, thereby reducing system energy consumption. Furthermore, by proactively controlling the heat source device's temperature, the temperature parameter threshold can be set more accurately, eliminating the need for empirical estimation. Furthermore, by determining the real-time control parameters of the heat sink based on the real-time temperature of the heat source device, determining target control parameters based on the optimized control parameters and the real-time control parameters for the future time period, and controlling the operation of the heat sink based on the target control parameters, the timely temperature control can be improved based on proactive temperature control of the heat source device, thereby addressing sudden temperature changes.

[0235] According to one embodiment of the present application, the determination module 430 is used to: adjust the weight value of the first optimized control quantity parameter and the weight value of the real-time control quantity parameter in the optimized control quantity parameter sequence according to the temperature gradient corresponding to the heat source device, and perform weighted calculation on the first optimized control quantity parameter and the real-time control quantity parameter to obtain the target control quantity parameter.

[0236] According to an embodiment of the present application, the determination module 430 is used to: adjust the weight value of the first optimized control quantity parameter, the weight value of the real-time control quantity parameter and the weight value of the feedforward control quantity parameter according to the optimized temperature parameter sequence corresponding to the optimized control quantity parameter sequence, the residual between the real-time temperature of the heat source device and the optimized temperature parameter in the optimized temperature parameter sequence, and at least one of the standard deviations of the residuals, and the temperature gradient corresponding to the heat source device, and perform weighted calculation on the first optimized control quantity parameter, the real-time control quantity parameter and the feedforward control quantity parameter to obtain the target control quantity parameter, wherein the feedforward control quantity parameter is determined based on the heat consumption of the heat source device.

[0237] According to an embodiment of the present application, when the temperature gradient corresponding to the heat source device is greater than or equal to a temperature gradient threshold, the robot in which the heat source device is located is in a transient operating state; when the temperature gradient corresponding to the heat source device is less than the temperature gradient threshold and the residual between the real-time temperature of the heat source device and the optimized temperature parameter is less than or equal to a first residual threshold, the robot is in a steady-state operating state; when the optimized temperature parameter is greater than or equal to a safety threshold, the robot is in a safe operating state; when the standard deviation of the residual is greater than or equal to a standard deviation threshold, the robot is in a parameter mismatch operating state, which is used to indicate that the parameters in the cost function do not match the current state of the robot. The weight value of the feedforward control quantity parameter in the transient operating state is greater than the weight value of the feedforward control quantity parameter in the steady-state operating state, the safe operating state, and the parameter mismatch operating state; the weight value of the first optimized control quantity parameter in the safe operating state is greater than the weight value of the first optimized control quantity parameter in the transient operating state, the steady-state operating state, and the parameter mismatch operating state; and the weight value of the real-time control quantity parameter in the parameter mismatch operating state is greater than the weight value of the real-time control quantity parameter in the transient operating state, the steady-state operating state, and the safe operating state.

[0238] According to one embodiment of the present application, the optimized control quantity parameter sequence corresponds to an optimized temperature parameter sequence, the optimized temperature parameter sequence includes the optimized temperature parameters of the heat source device at multiple moments in a future time period, the heat transfer process of the heat source device corresponds to a thermal resistance-heat capacitance equivalent circuit, and the temperature parameter sequence is determined based on the thermal resistance values ​​of multiple thermal resistors in the thermal resistance-heat capacitance equivalent circuit. The temperature control device 400 also includes a calculation module 450 for: calculating the real-time temperature gradient of the heat source device at a historical moment, and the optimized temperature gradient corresponding to the optimized temperature parameters of the heat source device at the historical moment; obtaining a temperature gradient residual based on the real-time temperature gradient and the optimized temperature gradient; and updating at least some of the thermal resistance values ​​in the cost function when the temperature gradient residual is greater than or equal to a second residual threshold.

[0239] According to one embodiment of the present application, the calculation module 450 is used to: obtain state parameters, where the state parameters include temperature parameters of the heat source device, ambient temperature parameters, and at least part of the thermal resistance value; and use the Kalman filtering method to update part of the thermal resistance value based on the state parameters.

[0240] According to one embodiment of the present application, the temperature control device 400 also includes an adjustment module 460, which is used to: when the gap between the preset temperature parameter threshold sequence and the optimized temperature parameter sequence corresponding to the optimized control parameter sequence meets the preset conditions, adjust the motion instructions of the robot where the heat source device is located, and the load of the joint corresponding to the heat source device in the adjusted motion instructions is less than the load of the joint corresponding to the heat source device in the motion instructions before adjustment.

[0241] According to one embodiment of the present application, the robot includes multiple joints, and the joint corresponding to the heat source device includes the first joint among the multiple joints, wherein the adjustment module 460 is used to: reduce the load of the first joint and increase the load of the second joint among the multiple joints.

[0242] According to one embodiment of the present application, the load of the first joint is negatively correlated with the temperature difference variable, wherein the temperature difference variable is used to characterize the ratio between the difference between the real-time temperature of the heat source device and the safety temperature corresponding to the heat source device and the difference between the warning temperature corresponding to the heat source device and the safety temperature corresponding to the heat source device.

[0243] It should be understood that the operations and functions of the optimization module 410, acquisition module 420, determination module 430, control module 440, calculation module 450 and adjustment module 460 in the above embodiment can refer to the above Figure 2 or Figure 3 In order to avoid repetition, the description of the temperature control method provided in the embodiment will not be repeated here.

[0244] Figure 5 FIG2 is a block diagram of an electronic device 500 for executing a temperature control method according to an exemplary embodiment of the present application. The electronic device 500 may be a server, a robot, a server interacting with a robot, or other devices.

[0245] Reference Figure 5 The electronic device 500 includes a processing component 510, which further includes one or more processors, and a memory resource represented by a memory 520 for storing instructions executable by the processing component 510, such as an application. The application stored in the memory 520 may include one or more modules, each corresponding to a set of instructions. In addition, the processing component 510 is configured to execute the instructions to perform the above-described temperature control method.

[0246] The electronic device 500 may further include a power supply component configured to perform power management of the electronic device 500, a wired or wireless network interface configured to connect the electronic device 500 to a network, and an input / output (I / O) interface. The electronic device 500 may be operated based on an operating system stored in the memory 520, such as Windows Server 2008. TM , Mac OSX TM , Unix TM , Linux TM , FreeBSD TM or similar.

[0247] A non-transitory computer-readable storage medium, when instructions in the storage medium are executed by a processor of the electronic device 500, enables the electronic device 500 to perform a temperature control method.

[0248] A computer program product includes a computer program. When the computer program is executed by a processor of a computer device, the computer device is enabled to execute the temperature control method provided in any of the above embodiments.

[0249] All of the above optional technical solutions can be arbitrarily combined to form optional embodiments of the present application, and will not be described in detail here.

[0250] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0251] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0252] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0253] The units described as separate components may or may not be physically separate, and 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 these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0254] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0255] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program check codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0256] It should be noted that, in the description of this application, the terms "first," "second," "third," etc. are used for descriptive purposes only and should not be understood as indicating or implying relative importance. In addition, in the description of this application, unless otherwise specified, "plurality" means two or more.

[0257] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of relevant countries and regions, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0258] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent replacements, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A temperature control method, characterized in that: include: optimizing a control quantity parameter sequence of a heat dissipation device in a future time period based on a cost function to obtain an optimized control quantity parameter sequence, wherein the heat dissipation device is used to dissipate heat from a heat source device, the cost function is constructed based on a difference between a temperature parameter sequence of the heat source device in the future time period and a preset temperature parameter threshold sequence, and a control quantity parameter sequence of the heat dissipation device, the temperature parameter sequence including temperature parameters of the heat source device at multiple moments in the future time period, and the value of the control quantity parameter sequence affecting the value of the temperature parameter sequence; Acquiring the real-time temperature of the heat source device to obtain a real-time temperature parameter sequence, and determining the real-time control quantity parameter of the heat dissipation device according to the real-time temperature parameter sequence and the preset temperature parameter threshold sequence; Determining a target control quantity parameter based on the optimized control quantity parameter sequence and the real-time control quantity parameter; The heat dissipation device is controlled to operate based on the target control parameter to achieve temperature control of the heat source device. The step of determining the target control quantity parameter based on the optimized control quantity parameter sequence and the real-time control quantity parameter includes: According to the temperature gradient corresponding to the heat source device, the weight value of the first optimized control parameter in the optimized control parameter sequence and the weight value of the real-time control parameter are adjusted, and the first optimized control parameter and the real-time control parameter are weightedly calculated to obtain the target control parameter.

2. The temperature control method according to claim 1, characterized in that: The step of adjusting the weight value of the first optimized control parameter and the weight value of the real-time control parameter in the optimized control parameter sequence according to the temperature gradient corresponding to the heat source device, and performing weighted calculation on the first optimized control parameter and the real-time control parameter to obtain the target control parameter includes: According to the optimized temperature parameter sequence corresponding to the optimized control parameter sequence, the residual between the real-time temperature of the heat source device and the optimized temperature parameter in the optimized temperature parameter sequence, and at least one of the standard deviations of the residuals, and the temperature gradient corresponding to the heat source device, the weight value of the first optimized control parameter, the weight value of the real-time control parameter and the weight value of the feedforward control parameter are adjusted, and the first optimized control parameter, the real-time control parameter and the feedforward control parameter are weightedly calculated to obtain the target control parameter, wherein the feedforward control parameter is determined based on the heat consumption of the heat source device.

3. The temperature control method according to claim 2, characterized in that: When the temperature gradient corresponding to the heat source device is greater than or equal to a temperature gradient threshold, the robot where the heat source device is located is in a transient working state; When the temperature gradient corresponding to the heat source device is less than the temperature gradient threshold and the residual between the real-time temperature of the heat source device and the optimized temperature parameter is less than or equal to a first residual threshold, the robot is in a steady-state working condition; When the optimized temperature parameter is greater than or equal to a safety threshold, the robot is in a safe working condition; When the standard deviation of the residual is greater than or equal to a standard deviation threshold, the robot is in a parameter mismatch condition, where the parameter mismatch condition is used to indicate that the parameters in the cost function do not match the current state of the robot. The weight value of the feedforward control parameter under the transient operating condition is greater than the weight value of the feedforward control parameter under the steady-state operating condition, the safe operating condition, and the parameter mismatch operating condition. The weight value of the first optimized control parameter under the safe operating condition is greater than the weight values ​​of the first optimized control parameter under the transient operating condition, the steady-state operating condition, and the parameter mismatch operating condition; The weight value of the real-time control parameter under the parameter mismatch condition is greater than the weight value of the real-time control parameter under the transient condition, the steady-state condition, and the safe condition.

4. The temperature control method according to claim 1, wherein: The optimized control quantity parameter sequence corresponds to an optimized temperature parameter sequence, the optimized temperature parameter sequence includes optimized temperature parameters of the heat source device at multiple moments in the future time period, the heat transfer process of the heat source device corresponds to a thermal resistance-heat capacitance equivalent circuit, and the temperature parameter sequence is determined based on the thermal resistance values ​​of multiple thermal resistors in the thermal resistance-heat capacitance equivalent circuit. The temperature control method further includes: Calculating the real-time temperature gradient of the heat source device at a historical moment, and the optimized temperature gradient corresponding to the optimized temperature parameter of the heat source device at the historical moment; Obtaining a temperature gradient residual according to the real-time temperature gradient and the optimized temperature gradient; When the temperature gradient residual is greater than or equal to a second residual threshold, at least part of the thermal resistance values ​​in the cost function are updated.

5. The temperature control method according to claim 4, characterized in that: The updating of at least part of the thermal resistance values ​​in the cost function includes: Acquiring state parameters, wherein the state parameters include temperature parameters of the heat source device, ambient temperature parameters, and the at least partial thermal resistance value; The partial thermal resistance value is updated based on the state parameter using a Kalman filter method.

6. The temperature control method according to any one of claims 1 to 5, characterized in that: Also includes: When the gap between the preset temperature parameter threshold sequence and the optimized temperature parameter sequence corresponding to the optimized control parameter sequence meets the preset conditions, the motion instruction of the robot where the heat source device is located is adjusted, and the load of the joint corresponding to the heat source device in the adjusted motion instruction is less than the load of the joint corresponding to the heat source device in the motion instruction before adjustment.

7. The temperature control method according to claim 6, characterized in that: The robot includes a plurality of joints, and the joint corresponding to the heat source device includes a first joint among the plurality of joints, wherein adjusting the motion instruction of the robot where the heat source device is located includes: The load on the first joint is reduced, and the load on the second joint among the plurality of joints is increased.

8. The temperature control method according to claim 7, characterized in that: The load of the first joint is negatively correlated with the temperature difference variable, wherein the temperature difference variable is used to characterize the ratio between the difference between the real-time temperature of the heat source device and the safety temperature corresponding to the heat source device and the difference between the warning temperature corresponding to the heat source device and the safety temperature corresponding to the heat source device.

9. A robot, characterized in that: The device comprises a control module, wherein the control module is used to execute the temperature control method according to any one of claims 1 to 8.

10. A computer program product, characterized in that The computer program product includes a computer program, and when the computer program is executed by a processor of a computer device, the computer device is enabled to execute the temperature control method according to any one of claims 1 to 8.

Images