Self-adaptive power distribution method for heating bionic skin of humanoid robot

Abstract

The invention belongs to the field of robot control, particularly relates to a bionic skin heating self-adaptive power distribution method for a humanoid robot, and aims to solve the problem that reasonable heating control is difficult to carry out on the humanoid robot when the humanoid robot executes a complex dynamic task in the prior art. The method comprises the following steps: acquiring target data; inputting the target data into a pre-constructed digital twinborn model to predict a thermal state change trend of each skin area in a future preset period, and identifying a key area which is heated preferentially in combination with the predictive motion state; taking the heating power distributed to each skin area as a decision variable, and constructing and solving a multi-objective optimization model; and determining a heating control instruction of each skin area according to the solved power distribution value. According to the method, the digital twinborn model and the predictive motion state are utilized, the thermal state trend caused by future actions and environment changes of the robot is predicted in advance, and the heating control effect of the robot under the complex working condition is effectively improved.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of robot control, and particularly relates to a human-shaped robot bionic skin heating adaptive power distribution method. BACKGROUND

[0002] Human-shaped robots are increasingly deployed in complex and variable unstructured environments such as outdoor patrol, disaster rescue, and human-robot collaboration. The bionic skin on the surface of the human-shaped robot is a key medium for interaction with the environment, and its temperature stability is crucial for the accuracy of the embedded sensors, the physical properties of the bionic materials, and the safety of human-robot interaction.

[0003] Existing robot heating control technologies, whether based on threshold PID (proportional-integral-derivative) control or preset fixed power levels, essentially simplify thermal management as a local, reactive temperature maintenance problem. These methods can cope with simple working conditions when the robot is static or the environment is stable, but when the human-shaped robot performs complex dynamic tasks, its inherent limitations are exposed. For example, the movement of the robot (such as high-speed swinging of the arm) will cause a dramatic and uneven change in the convective heat transfer coefficient of each part of the surface; its task (such as grabbing cold tools) will introduce strong local contact heat conduction, resulting in a high degree of spatiotemporal heterogeneity and dynamic change in the heat demand of the robot surface. Therefore, reactive control cannot compensate effectively before a large amount of heat is lost due to inherent hysteresis, and static power distribution strategies cannot adapt to dynamic changes in heat demand, resulting in waste in low-demand areas and insufficient heating in high-demand areas. SUMMARY

[0004] To solve the above problems in the prior art, i.e., the problem that related technologies cannot reasonably control the heating of human-shaped robots when they perform complex dynamic tasks, the present application proposes a human-shaped robot bionic skin heating adaptive power distribution method, which comprises: acquiring target data, including the surface temperature of each skin region of the robot, the environmental parameters of the environment, and the real-time and predictive motion state of the robot; inputting the target data into a pre-constructed digital twin model to predict the thermal state change trend of each skin region in a future preset period, and identifying the key regions that need to be heated preferentially in combination with the predictive motion state; A multi-objective optimization model is constructed and solved, taking the heating power distributed to each skin region as a decision variable; wherein the multi-objective optimization model takes the power distribution value of each skin region as a decision variable, takes the temperature of the key region being not lower than a preset threshold as a core constraint, and takes the minimization of total energy consumption and the maximization of temperature field uniformity of the overall surface of the robot as optimization objectives, with the optimization constraints being the upper limit of the total power of each skin region and the power limit value of the heating unit corresponding to each skin region; According to the obtained power distribution value, a heating control instruction for each skin region is determined.

[0005] In some preferred embodiments, the target data acquisition includes: Real-time temperature readings of N independent regions on the surface of the robot are concurrently collected at a preset collection frequency by temperature sensors integrated in advance under each skin region, to form an N-dimensional temperature state vector, where N is a positive integer; Temperature, relative humidity, and three-axis air flow rate data of the environment are collected to form an environmental parameter vector; A uniform global timestamp is added to the temperature state vector and the environmental parameter vector, encapsulated into a data structure body satisfying a preset format, and the data structure body is transmitted as input to a computing unit running a digital twin model.

[0006] In some preferred embodiments, the target data acquisition further includes: A main motion controller of the robot is accessed to extract a predetermined task trajectory within T seconds in the future, which includes target angle, angular velocity, and angular acceleration instructions in the joint space of the robot; A forward kinematics model of the robot established based on D-H parameters is used to perform frame-by-frame operation on the predetermined task trajectory, to convert it into time series data of three-dimensional coordinates, linear velocity, and linear acceleration of M control points preset on the surface of the robot's bionic skin in the world coordinate system, which is taken as the predictive motion state of the robot.

[0007] In some preferred embodiments, the identification of the key region to be preferentially heated in combination with the predictive motion state includes: According to the predictive motion state, the maximum linear velocity of each skin region control point within a preset period in the future is determined, and a region with a maximum linear velocity exceeding a first preset threshold is marked as a candidate key region; According to the predictive motion state, a skin region that will come into contact with an object with a temperature lower than a second preset threshold within a preset period in the future is determined and added to the candidate key region; A logical union operation is performed on each skin region in the candidate key region to determine the key region.

[0008] In some preferred embodiments, the digital twin model is a three-dimensional transient thermal finite element model of a multi-layer structure, and the prediction of the thermal state change trend of each skin region in the future preset period comprises: The surface temperature data of each skin region is used as the initial temperature field condition of the digital twin model, and the environmental parameters are used to define the external boundary condition of the model; The heat generated by the internal joint movement of the robot is used as an internal heat source, and a forward integral calculation is performed in the time domain by solving a non-steady-state heat conduction partial differential equation to output the predicted sequence of temperature change over time for each grid node in the future preset period.

[0009] In some preferred embodiments, the temperature field uniformity of the overall surface of the robot is maximized, and the implementation method comprises: In the objective function of the multi-objective optimization model, an unevenness evaluation function is defined, which is used to calculate the overall variance of the predicted temperature value set of all N skin regions at the end of the prediction period; The minimization of the unevenness evaluation function is used as the optimization target of the multi-objective optimization model, which is processed in parallel with the minimization of the total energy consumption; The power allocation decision variable is determined, which makes the two-dimensional target vector composed of the minimization of the unevenness evaluation function and the minimization of the total energy consumption reach the Pareto optimum under the premise of meeting the core constraints.

[0010] In some preferred embodiments, the multi-objective optimization model comprises: A population is randomly initialized, wherein each individual is a real-coded chromosome containing N skin region heating power allocation values; For each individual in the population, the temperature field is predicted according to its corresponding power allocation value through the digital twin model, and the values of the two objective functions, total energy consumption and temperature field variance, are calculated; Based on the calculated objective function values, the population is non-dominantly sorted and crowdedness is calculated, and according to the sorting results and crowdedness, the population is iterated through a polynomial mutation genetic operator to generate a child population; The parent and child populations are combined, and the above steps are repeated until the preset number of iterations or convergence conditions are met, and the optimal solution set of the current model is output.

[0011] In some preferred embodiments, the heating control instructions for each skin region are determined according to the power allocation values obtained by solving, comprising: From the optimal solution set obtained by solving, one of the power allocation vectors is selected as the execution scheme of the current control period according to the preset preference strategy; convert each determined power allocation value into a duty cycle of a pulse width modulation signal according to the electrical characteristics of the heating unit corresponding to the determined power allocation value; distribute the calculated N duty cycle values to the corresponding N skin area heating units respectively.

[0012] In some preferred embodiments, the preset threshold of the critical area includes the following steps: at the beginning of each control cycle, querying the task identifier of the task currently being executed or about to be executed from the task scheduling system of the robot; using the queried task identifier as a key, retrieving the corresponding temperature lower limit structure from the preset task-threshold mapping table, and extracting the values corresponding to each critical area as the preset threshold, the task-threshold mapping table is a key-value data structure, wherein the key is the task identifier of the robot, and the value is the temperature lower limit structure corresponding to different parts of the robot.

[0013] Advantages of the present application: The present application uses digital twin model and predictive motion state to predict the thermal state trend caused by future robot action and environmental changes, so that the power allocation decision based on this is no longer based on the temperature deviation that has occurred, but on the accurate prediction of future heat demand. Furthermore, the present application effectively overcomes the response lag problem of traditional control methods, ensures that the temperature of the critical area remains stable even under severe dynamic disturbance, and effectively improves the heating control effect of the robot under complex working conditions.

[0014] At the same time, the present application constructs a multi-objective optimization model for thermal management problem, which contains core constraints and double optimization objectives (total energy consumption, temperature uniformity), by solving its optimal set, a series of trade-off schemes are provided, solving the problem that multiple conflicting performance indicators are difficult to consider. BRIEF DESCRIPTION OF DRAWINGS

[0015] Other features, objects and advantages of the present application will become more apparent from the following detailed description of non-limiting embodiments made with reference to the accompanying drawings: Figure 1 is a flowchart of a human-shaped robot bionic skin heating adaptive power allocation method according to an embodiment of the present application; Figure 2 is a framework diagram of a human-shaped robot bionic skin heating adaptive power allocation system according to an embodiment of the present application Figure 3 is a structural diagram of a computer system according to an embodiment of the present application. DETAILED DESCRIPTION

[0016] The application will be described in further detail below with reference to the drawings and embodiments. It is to be understood that the specific embodiments described herein are intended to be illustrative only and not limiting of the application. It is also to be understood that the terminology used herein is for the purpose of describing the particular embodiments only and is not intended to be limiting.

[0017] It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0018] Please refer to Figure 1 The first embodiment of the present application provides a human-shaped robot bionic skin heating adaptive power distribution method, comprising: Step S10, obtaining target data, the target data including the surface temperature of each skin area of the robot, the environmental parameters of the environment, and the real-time and predictive motion state of the robot; Step S20, inputting the target data into a pre-constructed digital twin model to predict the thermal state change trend of each skin area in a future preset period, and identifying the key area to be preferentially heated in combination with the predictive motion state; Step S30, taking the heating power distributed to each skin area as a decision variable, constructing and solving a multi-objective optimization model; wherein the multi-objective optimization model takes the power distribution value of each skin area as a decision variable, takes the temperature of the key area being not lower than a preset threshold as a core constraint, and takes the minimization of total energy consumption and the maximization of temperature field uniformity of the overall surface of the robot as optimization objectives, and the optimization constraint is the upper limit of the total power of each skin area and the power limit value of the corresponding heating unit of each skin area; Step S40, determining the heating control instruction of each skin area according to the obtained power distribution value.

[0019] The method proposed in this embodiment can be executed by a human-shaped robot with bionic skin and capable of adaptive heating of the bionic skin and controlling the heating power thereof.

[0020] For example, the method is used for heating of bionic skin on a six-degree-of-freedom collaborative robot arm, the robot arm is deployed in an automated warehouse environment with an environmental temperature of 5℃ and irregular air flow, and the task of the robot arm is to identify and grasp metal and plastic parts of different materials. The bionic skin of the robot arm is divided into 50 independent heating areas (N=50), and each area is equipped with an independent thin-film resistance heating unit and a micro-electro-mechanical system (MEMS) temperature sensor.

[0021] The specific process of the method in this embodiment will be described below in combination with the robot described above.

[0022] Specifically, the method can be specifically executed by the control system of the robot, which first executes the data acquisition operation in a control cycle of 20 milliseconds.

[0023] Exemplarily, through the MEMS temperature sensor array deployed under the 50-skin-area silicone substrate, real-time temperature readings of each area are concurrently collected at a sampling frequency of 50 Hz, and the readings form a 50-dimensional real-time temperature vector; At the same time, an environment perception module installed on the robot base is called, which integrates a high-precision digital temperature and humidity sensor and a three-axis ultrasonic anemometer to obtain the current environmental temperature, relative humidity H_env and three-dimensional air flow velocity vector in real time. At the same time, the current angles and angular velocities of all joints are directly read from the robot motion controller. For example, by accessing the task planner of the robot, the planned trajectory within the next 5 seconds (i.e., the preset period T=5s) can be extracted. The trajectory is given in the form of a series of time stamp-joint angle target points. Using the established robot forward kinematics model (based on the D-H parameter method), the joint space trajectory is calculated into the time series of three-dimensional spatial positions, linear velocities and linear accelerations of the center points of each skin area in the world coordinate system.

[0024] In this embodiment, the pre-constructed digital twin model is a three-dimensional transient thermodynamic model based on the finite element method (FEM). The model is structurally identical to the physical entity of the robot and includes the accurate geometry and material thermal physical parameters (thermal conductivity, specific heat capacity, density, etc.) of the skin layer, heating layer, thermal insulation layer and internal skeleton.

[0025] At the beginning of each control cycle, the real-time temperature vector obtained in the foregoing steps is taken as the initial temperature field condition of the finite element model, and the environmental parameters and the predicted relative velocity sequence are taken as the boundary conditions of the model. The model solves the unsteady heat conduction equation and performs forward integration in the time domain to output the temperature change trend curves of all skin areas within the next 5 seconds.

[0026] At the same time, in combination with the predictive motion state described above, key areas in the multiple skin areas are judged and identified, including: By analyzing the speed prediction sequence within the next 5 seconds, any skin area with a predicted peak linear velocity exceeding 0.8 m / s (for example, the end of the arm when swinging quickly) is marked as a candidate key area; By analyzing the task instruction, for example, when a grasping instruction is identified, according to the target object attributes identified by the vision system, if the target is a metal part (whose temperature is by default close to the environmental temperature, i.e., 5℃), the skin areas of the palm and fingertips performing the grasping task are marked as candidate key areas; All the marked candidate areas are merged to form a final key area list.

[0027] On the basis of the above steps, a multi-objective optimization model is constructed and solved, specifically including: Determine the decision variable, i.e. the heating power allocation value of the 50 skin areas, to form a vector P; For all the areas i identified in the key area list, the minimum predicted temperature in the next 5-second prediction period must not be lower than the preset threshold (20℃ in this embodiment); The optimization objectives include: Objective 1 (energy consumption): minimize total heating power; Objective 2 (uniformity): maximize temperature field uniformity, which is equivalent to minimizing the variance of the predicted temperatures of all 50 areas at the end of the prediction period; The optimization constraints include: Total power upper limit: for example, the maximum available heating power provided by the robot battery management system is 200 watts; Individual power limit: for example, the rated maximum power of each heating unit is 8 watts.

[0028] Exemplarily, a non-dominated sorting genetic algorithm with elitism (NSGA-II) can be used to solve the multi-objective optimization problem. Each individual in the population of the algorithm is a 50-dimensional power allocation vector P. In the iteration process, for each individual, the digital twin model is called as an evaluator of the fitness function to calculate its corresponding total energy consumption and temperature variance. After selection, crossover, mutation, etc., a set of Pareto optimal solutions is finally converged and output. This set of solutions represents the best power allocation schemes with different trade-offs between energy consumption and uniformity while meeting all the constraints.

[0029] As a feasible implementation, taking the Pareto optimal solution set as an example, a final execution solution is selected according to the current working mode of the robot. For example: Energy-saving cruise mode: select the solution with the lowest total energy consumption in the solution set; Human-machine interaction mode: select the solution with the smallest temperature variance in the solution set; Standard operation mode: select the inflection point solution on the Pareto frontier, i.e. the solution with the most balanced energy consumption and uniformity.

[0030] Assuming that the optimal power allocation vector is selected in the above process, the system converts each power value corresponding to it into a duty cycle of the corresponding pulse width modulation (PWM) signal; Subsequently, the main controller transmits the 50 duty cycle values to 50 slave microcontrollers (MCUs) that control the heating units of the corresponding skin areas through the controller area network (CAN) bus. After receiving the instructions, each MCU immediately generates a PWM waveform with the specified duty cycle to accurately drive the connected thin-film heater.

[0031] In the above-mentioned embodiments, by cyclically executing the above-mentioned steps, the robot can be enabled to proactively and preferentially guarantee the temperature stability of the key parts in the most energy-saving manner in a dynamic and complex environment, while taking into account the overall surface thermal uniformity.

[0032] Further, in the above-mentioned embodiments, obtaining target data comprises: Concurrently collecting real-time temperature readings of N independent regions on the surface of the robot at a preset acquisition frequency through temperature sensors integrated in advance under each skin region, to form an N-dimensional temperature state vector, N being a positive integer; collecting temperature, relative humidity, and three-axis air flow rate data of the environment to form an environmental parameter vector; appending a unified global timestamp to the temperature state vector and the environmental parameter vector, encapsulating them into a data structure body satisfying a preset format, and transmitting the data structure body as input to a computing unit running a digital twin model.

[0033] The data structure body is used to encapsulate the thermal state information of the robot, and the sensor data collected in a dispersed manner is aggregated and stamped with a unified timestamp in a standardized manner, to ensure that the data input to the digital twin model has time synchronization and atomicity.

[0034] It is easy to understand that the data structure body logically includes three core components: Global timestamp: records the precise time when the data snapshot is created; Environmental parameters: encapsulate all thermodynamic parameters describing the external environment in which the robot is located.

[0035] Skin temperature state: an embedded sub-structure body for encapsulating temperature readings of all independent heating regions on the surface of the robot.

[0036] Further, in the above-mentioned embodiments, the obtaining target data further comprises: Accessing the main motion controller of the robot to extract a predetermined task trajectory within T seconds in the future, the predetermined task trajectory including target angle, angular velocity, and angular acceleration instructions in the joint space of the robot; using a robot forward kinematics model established based on D-H parameters, performing frame-by-frame operations on the predetermined task trajectory to convert it into time series data of three-dimensional coordinates, linear velocity, and linear acceleration of M control points preset on the surface of the robot's biomimetic skin in the world coordinate system, and using the time series data as the predicted motion state of the robot.

[0037] Specifically, taking the six-degree-of-freedom robot in the above-mentioned embodiments as an example, a standard Denavit-Hartenberg (D-H) parameter table is established according to its mechanical structure manual.

[0038] Exemplarily, the table contains six rows, each of which corresponds to a joint, and the columns contain four parameters of the link twist angle, the link length, the link offset, and the joint angle. Based on this D-H table, the forward kinematics function of the robot is constructed, in which the input is a 6-dimensional joint angle vector, and the output is a 4x4 homogeneous transformation matrix, which describes the pose of the robot end effector (the sixth link) relative to the robot base (the world coordinate system).

[0039] Meanwhile, through the communication interface of the robot operating system, the message published by the main motion controller is subscribed, which is used to indicate the scheduled task trajectory in the future period (in this embodiment, the preset period T = 5 seconds), and the trajectory data is organized as a series of waypoints with timestamps.

[0040] Exemplarily, assuming that 500 waypoints are extracted, batch operations are performed on the 500 extracted waypoints frame by frame (i.e., time point by time point). For each time point t in the trajectory: First, the pose of each link coordinate system is calculated, and the homogeneous transformation matrix from the base to each link is calculated according to the joint angle of the current frame by using the forward kinematics model; For the i-th control point (attached to the i-th link), its three-dimensional coordinates in the world coordinate system are calculated by the following formula: ; Where is the constant coordinate vector of the control point in the local coordinate system of the link i to which it belongs; Then the linear velocity of each control point is calculated by using the geometric Jacobian matrix of the robot: ; Where, is the velocity component part of the Jacobian matrix related to the control point Pⱼ, is the joint angular velocity vector of the current frame. The Jacobian matrix itself is also a function of the joint angle .

[0041] The formula for calculating the linear acceleration of the control point is: ; Where is the derivative of the Jacobian matrix with respect to time, which depends on and ; is the joint angular acceleration vector of the current frame. ​

[0042] Further, in the above embodiment, in combination with the predicted motion state, the key region to be preferentially heated is identified, including: According to the predicted motion state, the maximum linear speed of each skin region control point in the future preset period is determined, and the region whose maximum linear speed exceeds a first preset threshold is marked as a candidate key region; according to the predicted motion state, the skin region that will be in contact with an object whose temperature is lower than a second preset threshold in the future preset period is determined and added to the candidate key region; a logical union operation is performed on each skin region in the candidate key region to determine the key region.

[0043] The first preset threshold is a maximum linear speed threshold, which is set to, for example, 1.2 m / s. The value can be obtained according to experimental data. When the motion speed of the skin region exceeds this value, the convective exchange effect with the ambient air will be significantly enhanced, resulting in rapid heat loss. The second preset threshold is a target object temperature threshold, which is set to, for example, 10.0 degrees Celsius. In a warehouse environment, an object (usually a metal part) with a temperature lower than this temperature will cause a sharp local temperature drop (conductive cooling) in the skin region in contact with it, affecting the tactile sensing accuracy of the robot.

[0044] Further, in the above embodiment, the digital twin model is a multi-layer three-dimensional transient thermodynamic finite element model, and the prediction of the thermal state change trend of each skin region in the future preset period includes: The surface temperature data of each skin region is used as the initial temperature field condition of the digital twin model, and the environmental parameters are used to define the external boundary condition of the model. The heat generated by the internal joint motion of the robot is used as an internal heat source. By solving the non-steady-state heat conduction partial differential equation, forward integration calculation is performed in the time domain, and the prediction sequence of the temperature change of each grid node over time in the future preset period is output.

[0045] The construction of the digital twin model can be based on the CAD model of the robot, and the accurate geometric shape of the outer skin and the internal support structure is extracted. For example, a mixed tetrahedral and hexahedral grid can be used to spatially discretize the geometric domain to generate a three-dimensional finite element grid. It includes at least two material layers: a biomimetic skin material layer on the surface and a structural support layer on the inside, and each layer is assigned its corresponding material thermal property parameters, including density, specific heat capacity, and anisotropic thermal conductivity.

[0046] The surface temperature data of each skin region collected by the distributed temperature sensor network at the current time is mapped to the corresponding surface nodes of the finite element model through an interpolation algorithm to serve as the initial condition for the entire three-dimensional temperature field transient simulation.

[0047] Meanwhile, according to the measured environmental wind speed and air temperature, the surface convection heat transfer coefficient is calculated according to an empirical formula, and is applied to all unit surfaces of the robot outer surface in contact with air. The specific empirical formula is not limited in the embodiment.

[0048] It is easy to understand that the change of the robot motion state will update the intensity and distribution of the internal heat source in real time. Therefore, the heat generated by the motion components such as the joint motors and reducers inside the robot is equivalent to the internal heat source in the volume, the power of the internal heat source can be determined by querying the mapping table of motor current, rotating speed and efficiency, and the power is distributed to the corresponding grid cells in the finite element model according to the spatial position.

[0049] After the above setting is completed, the three-dimensional unsteady heat conduction partial differential equation is solved, and the equation is numerically integrated, starting from the initial time, and gradually advancing the calculation at a fixed time step, simulating the temperature field evolution in a preset period (for example, 5-10 minutes in the future) in the future, and finally outputting the complete prediction sequence of the temperature change with time of each skin area in the time period.

[0050] Further, in the above embodiment, the temperature field uniformity of the whole surface of the robot is maximized, and the implementation method comprises: In the objective function of the multi-objective optimization model, an unevenness evaluation function is defined, the unevenness evaluation function is used to calculate the overall variance of the predicted temperature value set of all N skin areas at the end of the prediction period; the minimization of the unevenness evaluation function is taken as the optimization target of the multi-objective optimization model, and is processed in parallel with the minimization of the total energy consumption; the power distribution decision variable is determined, which makes the two-dimensional target vector composed of the minimization of the unevenness evaluation function and the minimization of the total energy consumption reach the Pareto optimality under the premise of meeting the core constraint.

[0051] Wherein, the unevenness evaluation function is defined as the overall variance of the predicted temperature value of all N skin areas at the end of the prediction time domain, and the minimization of the unevenness evaluation function is taken as an independent optimization target, and the original total energy consumption minimization target constitutes a two-dimensional target vector, which participates in optimization solving together with the core constraint condition in the optimization model.

[0052] For example, a comprehensive objective function can be constructed, wherein the weight coefficients respectively reflect the importance of the energy consumption target and the uniformity target, and a series of Pareto optimal solutions are generated by adjusting the ratio of the weight coefficients.

[0053] In each control cycle, a power distribution scheme that minimizes the comprehensive objective function is solved under the condition of meeting the total power constraint and the power limit of each region. A sequential quadratic programming algorithm can be used in the solving process to find the optimal power distribution decision variable that meets all the constraint conditions through iterative calculation. Finally, the obtained power distribution value is converted into the control instruction output of each heating unit.

[0054] Furthermore, in the above embodiment, the solving process of the multi-objective optimization model includes: Randomly initialize a population, where each individual is a real-coded chromosome containing N skin region heating power distribution values; for each individual in the population, predict the temperature field according to its corresponding power distribution value through the digital twin model, and calculate the total energy consumption and temperature field variance of two objective function values; based on the calculated objective function values, perform non-dominated sorting and crowding calculation on the population, and according to the sorting results and crowding, generate a child population through a polynomial mutation genetic operator; combine the parent and child populations, repeat the above steps until the preset iteration number or convergence condition is met, and output the optimal solution set of the current model.

[0055] In this embodiment, first set the population size to M, and each individual is represented by a real-coded chromosome, which represents a complete power distribution scheme, i.e. the chromosome form is [P1, P2,..., P_N], where P_i represents the heating power value of the i-th skin region. Randomly generate an initial population under the condition of meeting the system total power constraint and the power limit of each region.

[0056] Exemplarily, for each individual in the population, the following calculations are performed: Decode the chromosome into specific power distribution values; input this power distribution into the digital twin model for forward simulation to obtain the predicted temperature field distribution at the end of the time domain; calculate two objective function values: total energy consumption and temperature field unevenness; According to the objective function values, perform non-dominated sorting on the population, and divide the population into multiple front levels. The individuals in the first front are considered optimal, followed by the second front, and so on. In the same front, calculate the crowding degree of each individual to evaluate the distribution density of the individual in the target space.

[0057] On this basis, a binary tournament selection mechanism can be used to preferentially select individuals with high front level and large crowding degree as parents. Through the simulation of a binary crossover operator for gene recombination and a polynomial mutation operator for introducing perturbations, a child population is generated. Combine the parent and child populations to form a new population with a size of 2M.

[0058] The non-dominated sorting and congestion calculation are performed again on the combined population, and the first M individuals are selected as a new generation population according to the sorting result. The iteration process is repeated until the preset maximum iteration number G_max is reached, or the quality of the solution set is no longer significantly improved in continuous iterations, and finally all solutions of the first front are output as the Pareto optimal solution set.

[0059] Further, in the above embodiment, the preset threshold of the key area, the setting process includes: At the beginning of each control cycle, the task identifier currently being executed or about to be executed is queried from the task scheduling system of the robot; the task identifier queried is used as a key to retrieve the corresponding temperature lower limit structure in the preset task-threshold mapping table, and the values corresponding to each key area in the temperature lower limit structure are extracted as the preset threshold, the task-threshold mapping table is a key-value pair data structure, wherein the key is the task identifier of the robot, and the value is the temperature lower limit structure corresponding to different parts of the robot.

[0060] Specifically, at the beginning of each control cycle, the task identifier currently being executed is obtained in real time through the task scheduling system interface of the robot. The identifier adopts a unified task coding format and can uniquely identify the working state and task type of the robot.

[0061] In the embodiment, a task-threshold mapping table is established in advance, which is implemented in a hash table data structure for fast query. Each entry in the table contains two fields: a task identifier (key), a string type, which stores a standardized task name; and a temperature threshold structure (value), a structure type, which contains multiple subfields corresponding to the minimum temperature requirements of different key parts of the robot The task identifier obtained is used as a query key to perform a lookup operation in the task-threshold mapping table. If a matching entry is found, the corresponding temperature threshold structure is read; if no match is found, a default temperature threshold structure is used. The specific temperature threshold values of each key area are extracted from the retrieved structure.

[0062] The second embodiment of the present application proposes a human-shaped robot bionic skin heating adaptive power distribution system, comprising: The data acquisition module 210 is configured to acquire target data, wherein the target data includes surface temperatures of each skin area of the robot, environmental parameters of an environment in which the robot is located, and real-time and predictive motion states of the robot. The model prediction module 220 is configured to input the target data into a pre-constructed digital twin model to predict a thermal state change trend of each skin area in a future preset period, and identify a key area to be preferentially heated in combination with the predictive motion state. The model solving module 230 is configured to construct and solve a multi-objective optimization model by taking the heating power distributed to each skin region as a decision variable, wherein the multi-objective optimization model takes the power distribution value of each skin region as a decision variable, takes the temperature of the key region being not lower than a preset threshold as a core constraint, and takes the minimization of total energy consumption and the maximization of temperature field uniformity of the overall surface of the robot as optimization objectives, and the optimization constraint is the upper limit of the total power of each skin region and the power limit value of the heating unit corresponding to each skin region. The heating control module 240 is configured to determine the heating control instruction of each skin region according to the power distribution value obtained by solving.

[0063] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process and related description of the system described above can refer to the corresponding process in the foregoing method embodiments, which will not be described herein.

[0064] Reference will be made to the following description of the drawings. Figure 3 which shows the structural schematic diagram of a computer system of a server suitable for implementing the system and method embodiments of the present application. Figure 3 The server shown is only an example, and should not bring any limitation to the functions and use range of the embodiments of the present application.

[0065] As shown in Figure 3 , the computer system includes a central processing unit (CPU) 301, which can perform various appropriate actions and processes according to programs stored in a read-only memory (ROM) 302 or programs loaded from a storage portion 308 to a random access memory (RAM) 303. In the random access memory 303, various programs and data required for system operation are also stored. The central processing unit 301, the read-only memory 302, and the random access memory 303 are connected to each other through a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0066] The following components are connected to the input / output interface 305: an input section 306 including input devices such as a keyboard and a mouse; an output section 307 including output devices such as a cathode ray tube (CRT), a liquid crystal display (LCD), and a speaker; a storage section 308 including a hard disk; and a communication section 309 including a network interface card such as a LAN (Local Area Network) card, a modem, and the like. The communication section 309 performs communication processing via a network such as the Internet. A drive 310 is also connected to the input / output interface 305 as necessary. A removable media 311 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, and the like is attached to the drive 310 as necessary, so that a computer program read therefrom is installed in the storage section 308 as necessary.

[0067] In particular, the processes described above with reference to the flowcharts can be implemented as a computer software program according to embodiments of the present application. For example, embodiments of the present application include a computer program product comprising a computer program carried on a computer readable medium, the computer program containing program code for executing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via the communication section 309, and / or installed from the removable media 311. When the computer program is executed by the central processing unit 301, the above-described functions defined in the methods of the present application are performed. It should be noted that the above-described computer readable medium of the present application can be a computer readable signal medium or a computer readable storage medium or any combination of the two. The computer readable storage medium may, for example, be - but is not limited to - an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus or device, or any combination of the above.

[0068] More specific examples of the computer-readable storage medium can include but are not limited to the following: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the present disclosure, the computer-readable storage medium can be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. Also, in the present disclosure, the computer-readable storage medium can be any tangible medium that can be used to store or transfer a program for use by or in connection with an instruction execution system, apparatus, or device. The program code can be transmitted in a computer-readable signal medium using any suitable medium of modulation, including but not limited to wireless, wire line, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

[0069] Computer program code for carrying out operations of the present application can be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C++ or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider).

[0070] The computer program product of the present application can be a computer program implemented on one or more computers in one or more locations. The program instructions can be stored on a computer readable medium, such as a hard disk, CD-ROM, optical storage, or any other tangible medium. The program instructions can be downloaded over a network from a remote computer (e.g., a server computer) or to a remote computer (e.g., a client computer) with the assistance of one or more of the various above-mentioned network devices. The program instructions can be downloaded from the network or uploaded to the network.

[0071] The terms "first", "second", and the like, are used to distinguish between similar objects, not to denote a particular order or sequence.

[0072] The term "comprising" or any other similar term is intended to encompass the inclusion of one or more elements, not the exclusion of any other elements. In other words, the term "comprising" or any other similar term means that other elements can also be present.

[0073] So far, the technical solutions of the present application have been described in combination with the preferred embodiments shown in the drawings.

[0074] The above description is merely illustrative of the application, and is not intended to limit the application. The application can have various modifications and alterations without departing from the spirit and scope of the application. Any modifications, equivalent replacements, improvements, and the like within the spirit and principles of the application are intended to be included in the scope of the appended claims.

Claims

1. A method for human-shaped robot bionic skin heating adaptive power distribution, characterized in that, The method comprises: acquiring target data, the target data comprising surface temperatures of each skin region of a robot, environmental parameters of an environment in which the robot is located, and real-time and predictive motion states of the robot; inputting the target data into a pre-constructed digital twin model to predict thermal state change trends of each skin region in a future preset period, and identifying a key region to be preferentially heated in combination with the predictive motion states; constructing and solving a multi-objective optimization model with heating power allocated to each skin region as a decision variable; wherein the multi-objective optimization model takes the power allocation values of each skin region as a decision variable, takes the temperature of the key region being not lower than a preset threshold as a core constraint, and takes minimizing total energy consumption and maximizing temperature field uniformity of the overall surface of the robot as optimization objectives, with the optimization constraints being an upper limit of total power of each skin region and a power limit value of a corresponding heating unit of each skin region; determining heating control instructions for each skin region according to the power allocation values obtained by solving.

2. The human-robot bionic skin heating adaptive power allocation method according to claim 1, wherein, The acquisition of the target data comprises: concurrently collecting real-time temperature readings of N independent regions on the surface of the robot at a preset collection frequency through temperature sensors pre-integrated under each skin region, to form an N-dimensional temperature state vector, N being a positive integer; collecting temperature, relative humidity, and three-axis air flow rate data of the environment to form an environmental parameter vector; appending a unified global time stamp to the temperature state vector and the environmental parameter vector, encapsulating the data structure body satisfying the preset format, and transmitting the data structure body as input to a computing unit running the digital twin model.

3. The human-robot bionic skin heating adaptive power allocation method according to claim 2, wherein, The acquisition of the target data further comprises: accessing a main motion controller of the robot to extract a predetermined task trajectory in the next T seconds, the predetermined task trajectory comprising target angle, angular velocity, and angular acceleration instructions in the joint space of the robot; using a robot forward kinematics model established based on D-H parameters to perform frame-by-frame operations on the predetermined task trajectory, to convert the predetermined task trajectory into time series data of three-dimensional coordinates, linear velocity, and linear acceleration of M control points pre-positioned on the surface of the robot's bionic skin in the world coordinate system, and taking the time series data as the predictive motion state of the robot.

4. The human-robot bionic skin heating adaptive power allocation method of claim 1, wherein, The identification of the key region to be preferentially heated in combination with the predictive motion state comprises: determining maximum linear velocities of control points of each skin region in a future preset period according to the predictive motion state, and marking regions with maximum linear velocities exceeding a first preset threshold as candidate key regions; determining skin regions that will come into contact with objects having temperatures lower than a second preset threshold in the future preset period according to the predictive motion state, and adding the skin regions to the candidate key regions; performing a logical union operation on each skin region in the candidate key regions to determine the key region.

5. The human-robot bionic skin heating adaptive power allocation method of claim 1, wherein, The digital twin model is a multi-layer three-dimensional transient thermodynamic finite element model, and the prediction of thermal state change trends of each skin region in a future preset period comprises: taking surface temperature data of each skin region as an initial temperature field condition of the digital twin model, and using the environmental parameters to define external boundary conditions of the model; The heat generated by the internal joint movement of the robot is taken as an internal heat source, and a non-steady-state heat conduction partial differential equation is solved to perform forward integral calculation in the time domain, and a prediction sequence of temperature changes of each grid node over time in a future preset period is output.

6. The human-robot bionic skin heating adaptive power allocation method of claim 1, wherein, The method for maximizing the temperature field uniformity of the overall surface of the robot comprises the following steps: In the objective function of the multi-objective optimization model, an unevenness evaluation function is defined, which is used to calculate the overall variance of the predicted temperature value set of all N skin areas at the end of the prediction period; The minimization of the unevenness evaluation function is taken as the optimization target of the multi-objective optimization model, which is processed in parallel with the minimization of the total energy consumption; The power allocation decision variable is determined to make the two-dimensional target vector composed of the minimization of the unevenness evaluation function and the minimization of the total energy consumption reach the Pareto optimum under the premise of meeting the core constraints.

7. The human-robot bionic skin heating adaptive power allocation method of claim 1, wherein, The solving process of the multi-objective optimization model comprises the following steps: Randomly initialize a population, wherein each individual is a real-coded chromosome containing N skin area heating power allocation values; For each individual in the population, the temperature field is predicted based on the corresponding power allocation value through the digital twin model, and the total energy consumption and temperature field variance are calculated as two objective function values; Based on the calculated target function values, the population is non-dominantly sorted and crowdedness is calculated, and according to the sorting results and crowdedness, the population is iterated through the polynomial mutation genetic operator to generate a child population; Combine the parents and children, repeat the above steps until the preset iteration number or convergence condition is met, and output the optimal solution set of the current model.

8. The human-robot bionic skin heating adaptive power allocation method of claim 1, wherein, The heating control instructions of each skin area are determined according to the power allocation values obtained by solving, comprising: From the optimal solution set obtained by solving, one power allocation vector is selected as the execution scheme of the current control period according to the preset preference strategy; Each determined power allocation value is converted into a duty cycle of a pulse width modulation signal according to the electrical characteristics of the corresponding heating unit; The calculated N duty cycle values are respectively sent to the corresponding N skin area heating units.

9. The human-robot bionic skin heating adaptive power allocation method of claim 1, wherein, The preset threshold of the key area comprises the following steps: At the beginning of each control period, the task identifier currently being executed or about to be executed is queried from the task scheduling system of the robot; The task identifier obtained by querying is used as a key to retrieve the corresponding temperature lower limit structure in the preset task-threshold mapping table, and the values corresponding to each key area in the temperature lower limit structure are extracted as the preset threshold, wherein the task-threshold mapping table is a key-value data structure, and the key is the task identifier of the robot, and the value is the temperature lower limit structure corresponding to different parts of the robot.

Images