Hardware resource scheduling method and device of robot, electronic equipment and storage medium

Abstract

This application provides a method, apparatus, electronic device, and storage medium for scheduling hardware resources of a robot, belonging to the field of artificial intelligence technology and applicable to the fields of fintech and smart healthcare. The method includes: acquiring the robot's battery level data and the type of task to be performed, and acquiring the energy consumption load data of all hardware units. Then, energy consumption is evaluated based on the battery level data and energy consumption load data to obtain an energy consumption score. Next, a target unit is determined from multiple hardware units based on the battery level data and energy consumption load data. Subsequently, the task priority of the task type is acquired, and setting parameters are predicted based on the task priority, energy consumption score, and at least one of the target unit to obtain target setting parameters. Finally, hardware resources are scheduled for the hardware units based on the target setting parameters. This application embodiment can improve the energy utilization rate of the robot.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and is applicable to the fields of fintech and smart healthcare. In particular, it relates to a method and apparatus for scheduling hardware resources of a robot, an electronic device, and a storage medium. Background Technology

[0002] Currently, in the fintech field, robots can provide consulting services related to financial services, enabling efficient interaction with customers. In the field of smart healthcare, robots can be used to perform surgeries. These intelligent robots are typically deployed with large models to achieve the relevant functions. In inference applications, these large models usually only allocate computing resources according to the needs of the task, without considering the robot's energy consumption status (such as the remaining battery power). This can lead to situations where the robot's battery is insufficient to support the completion of high-energy-consuming tasks, or where it excessively consumes computing resources during low-load tasks, thus accelerating energy consumption and resulting in low energy efficiency. Therefore, how to improve the energy efficiency of robots has become an urgent technical problem to be solved. Summary of the Invention

[0003] The main objective of this application is to propose a method and apparatus for scheduling hardware resources of a robot, as well as an electronic device and storage medium, in order to improve the energy utilization rate of the robot.

[0004] To achieve the above objectives, a first aspect of this application proposes a hardware resource scheduling method for a robot, wherein the robot has at least two hardware units, and the method includes: Obtain the robot's battery power data and task type, and obtain the energy consumption load data of all hardware units; Based on the battery power data and the energy load data, the energy consumption level is evaluated to obtain an energy consumption score; The target unit is determined from the plurality of hardware units based on the battery power data and the energy load data; Obtain the task priority of the task type to be executed, and predict the setting parameters based on the task priority, the energy consumption score and at least one of the target units to obtain the target setting parameters; Hardware resource scheduling is performed on the hardware unit based on the target setting parameters.

[0005] In some embodiments, the step of predicting setting parameters based on at least one of the task priority, the energy consumption score, and the target unit to obtain target setting parameters includes: Obtain the working stage type of the robot; wherein, the working stage type includes the application stage and the model training stage; If the work phase type is the model training phase, the energy consumption score and the task priority will be used as reference data. If the work stage type is the application stage, the task priority and the hardware type of the target unit shall be used as reference data. Parameters are generated based on the reference data to obtain the target setting parameters for each hardware unit.

[0006] In some embodiments, when the reference data includes the task priority and the hardware type of the target unit, the step of generating parameters based on the reference data to obtain the target setting parameters for each hardware unit includes: The hardware type of the target unit is queried according to the preset hardware type parameter mapping relationship to obtain the target priority and the reference setting parameters matching the target priority. The execution task type with the task priority equal to the target priority is taken as the target task type, and a reference unit is determined from the hardware unit based on the target task type; The target setting parameters are determined based on the reference setting parameters of the reference unit and the original setting parameters of the original unit; wherein, the original unit is a hardware unit within the robot that is not the reference unit.

[0007] In some embodiments, when the reference data includes the energy consumption score and the task priority, the step of generating parameters based on the reference data to obtain the target setting parameters for each hardware unit includes: The parameter task mapping relationship is determined based on the energy consumption score and the score threshold. The target priority is determined from the task priorities based on the energy consumption score and the preset score threshold, and the target priority is mapped according to the parameter task mapping relationship to obtain the target setting parameters.

[0008] In some embodiments, the task priorities include high priority, medium priority, and low priority. The step of determining a target priority from the task priorities based on the energy consumption score and a preset score threshold, and mapping the target priority according to the parameter task mapping relationship to obtain the target setting parameters, includes: If the energy consumption score is less than the preset score threshold, the target priority includes high priority, medium priority and low priority; If the energy consumption score is greater than or equal to the preset score threshold, then the target priority includes low priority; The target priority is mapped according to the parameter task mapping relationship to obtain the target setting parameters.

[0009] In some embodiments, the hardware unit includes a battery, and determining the target unit from a plurality of hardware units based on the battery power data and the energy load data includes: Obtain the energy consumption parameter threshold matching each of the aforementioned hardware units, and obtain the battery power threshold. If the battery power data is less than the battery power threshold, then the battery is determined to be the target unit; For each of the aforementioned hardware units, if the energy consumption load data is greater than or equal to the energy consumption parameter threshold, then the hardware unit is determined to be the target unit.

[0010] In some embodiments, after performing hardware resource scheduling on the hardware unit based on the target setting parameters, the method further includes: The energy load data is updated to obtain intermediate energy consumption data; For each of the aforementioned hardware units, the energy consumption reduction rate is calculated based on the energy consumption load data and the intermediate energy consumption data; The performance data of the hardware unit is obtained, and the energy consumption reduction rate and the performance data are verified according to the preset scheduling iteration termination condition to obtain the verification result. If the verification result reflects that at least one of the energy consumption reduction rate and the performance data does not meet the scheduling iteration termination condition, the steps of predicting setting parameters based on at least one of the task priority, the energy consumption score and the target unit to obtain target setting parameters, and scheduling hardware resources for the hardware unit based on the target setting parameters are re-executed until both the energy consumption reduction rate and the performance data meet the scheduling iteration termination condition.

[0011] To achieve the above objectives, a second aspect of this application provides a hardware resource scheduling device for a robot, the device comprising: The acquisition module is used to acquire the robot's battery power data and the type of task to be performed, and to acquire the energy consumption load data of all the hardware units. The energy consumption assessment module is used to assess the energy consumption level based on the battery power data and the energy load data, and obtain an energy consumption score. A unit determination module is used to determine a target unit from a plurality of hardware units based on the battery power data and the energy consumption load data; The parameter prediction module is configured to obtain the task priority of the task type to be executed, and to predict the setting parameters based on at least one of the task priority, the energy consumption score and the target unit to obtain the target setting parameters. The hardware resource scheduling module is used to schedule hardware resources for the hardware unit based on the target setting parameters.

[0012] To achieve the above objectives, a third aspect of this application provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the method described in the first aspect.

[0013] To achieve the above objectives, a fourth aspect of the present application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method described in the first aspect.

[0014] The robot hardware resource scheduling method, apparatus, electronic device, and storage medium proposed in this application collect data on the robot's battery level, task type, and hardware unit energy consumption load. Then, energy consumption is assessed based on the battery level and energy consumption load data to obtain an energy consumption score. Next, target units are determined based on the battery level and energy consumption load data. Then, target setting parameters are determined based on at least one of the task priority of the task type, the energy consumption score, and the target unit. Finally, hardware resource scheduling is performed based on the target setting parameters. The method in this application effectively avoids situations where the robot cannot complete high-energy-consuming tasks due to insufficient power, and avoids unnecessary energy waste caused by excessive consumption of computing resources during low-load tasks. Ultimately, it improves the robot's energy utilization rate, not only extending the robot's endurance but also improving its operating efficiency, thereby reducing energy waste and enabling the robot to provide more durable and reliable services. Attached Figure Description

[0015] Figure 1 This is a flowchart of the hardware resource scheduling method for a robot provided in an embodiment of this application; Figure 2 yes Figure 1 The flowchart of step S103 in the process; Figure 3 yes Figure 1 The flowchart of step S104 in the process; Figure 4 yes Figure 3 The flowchart of step S304 in the process; Figure 5 yes Figure 3 Another flowchart of step S304 in the process; Figure 6 This is another flowchart of the robot hardware resource scheduling method provided in the embodiments of this application; Figure 7 This is a schematic diagram of the hardware resource scheduling device for the robot provided in the embodiments of this application; Figure 8This is a schematic diagram of the hardware structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0017] It should be noted that although functional modules are divided in the device schematic diagram and a logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the device or the order in the flowchart. The terms "first," "second," etc., in the specification, claims, and the aforementioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0019] First, let's analyze some of the terms used in this application: Artificial intelligence (AI) is a new branch of computer science that studies, develops, and applies theories, methods, technologies, and systems to simulate, extend, and expand human intelligence. It aims to understand the essence of intelligence and produce intelligent machines that can react in a way similar to human intelligence. Research in this field includes robotics, speech recognition, image recognition, natural language processing, and expert systems. AI can simulate the information processes of human consciousness and thought. Furthermore, AI utilizes digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceiving the environment, acquiring knowledge, and using that knowledge to achieve optimal results.

[0020] Currently, in the fintech field, robots can provide consulting services related to financial services, enabling efficient interaction with customers. In the field of smart healthcare, robots can be used to perform surgeries. These intelligent robots are typically deployed with large models to achieve the relevant functions. In inference applications, these large models usually only allocate computing resources according to the needs of the task, without considering the robot's energy consumption status (such as the remaining battery power). This can lead to situations where the robot's battery is insufficient to support the completion of high-energy-consuming tasks, or where it excessively consumes computing resources during low-load tasks, thus accelerating energy consumption and resulting in low energy efficiency. Therefore, how to improve the energy efficiency of robots has become an urgent technical problem to be solved.

[0021] Based on this, embodiments of this application provide a method and apparatus for scheduling hardware resources of a robot, an electronic device and a storage medium, aiming to improve the energy utilization rate of the robot.

[0022] The hardware resource scheduling method and apparatus, electronic device and storage medium for robots provided in this application are specifically described through the following embodiments. First, the hardware resource scheduling method for robots in this application is described.

[0023] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence (AI) refers to the theories, methods, technologies, and application systems that use digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0024] Foundational technologies for artificial intelligence generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interactive systems, and mechatronics. AI software technologies mainly encompass computer vision, robotics, biometrics, speech processing, natural language processing, and machine learning / deep learning.

[0025] The robot hardware resource scheduling method provided in this application relates to the field of artificial intelligence technology. This method can be applied to a terminal, a server, or software running on either a terminal or a server. In some embodiments, the terminal can be a smartphone, tablet, laptop, desktop computer, etc.; the server can be configured as an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms; the software can be an application implementing the robot hardware resource scheduling method, but is not limited to the above forms.

[0026] This application can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0027] Figure 1 This is an optional flowchart of the robot hardware resource scheduling method provided in the embodiments of this application. Figure 1 The method may include, but is not limited to, steps S101 to S105. It should be noted that the robot has at least two hardware units, which include a battery, processor, sensors, and actuators, etc.

[0028] Step S101: Obtain the robot's battery power data and task type, and obtain the energy load data of all hardware units.

[0029] Step S102: Evaluate the energy consumption level based on battery power data and energy load data to obtain an energy consumption score.

[0030] Step S103: Determine the target unit from multiple hardware units based on battery power data and energy load data.

[0031] Step S104: Obtain the task priority of the task type to be executed, and predict the setting parameters based on at least one of the task priority, energy consumption score and target unit to obtain the target setting parameters.

[0032] Step S105: Perform hardware resource scheduling on the hardware unit based on the target setting parameters.

[0033] Steps S101 to S105 of this embodiment involve collecting battery power data, task type, and hardware unit energy consumption load data of the robot. Energy consumption is then assessed based on the battery power and energy consumption load data to obtain an energy consumption score. Next, a target unit is determined based on the battery power and energy consumption load data. Then, target setting parameters are determined based on at least one of the task priority of the task type, the energy consumption score, and the target unit. Finally, hardware resource scheduling is performed based on the target setting parameters. The method of this embodiment effectively avoids situations where the robot cannot complete high-energy-consuming tasks due to insufficient power, and avoids unnecessary energy waste caused by excessive consumption of computing resources during low-load tasks. Ultimately, it improves the robot's energy utilization rate, not only extending its runtime but also increasing its operating efficiency, thereby reducing energy waste and enabling the robot to provide more durable and reliable services.

[0034] In step S101 of some embodiments, battery power data refers to the current remaining power of the robot battery, which is usually expressed in the form of voltage or power percentage. The battery voltage, current and temperature data can be monitored in real time by the battery management system (BMS) to calculate the current remaining battery power.

[0035] The type of task being performed refers to the type of task the robot is performing, such as navigation, object recognition, or voice interaction.

[0036] Hardware units refer to the various components of a robot, including batteries, processors, sensors, and actuators. Energy consumption load data refers to the energy consumption status data of each hardware unit, including the power consumption and load of each unit such as the processor, sensors, and actuators during the current task. The power consumption and load information of each unit are collected in real time through a hardware energy consumption detection module. For example, the energy consumption load data for the processor (CPU) is 20W, the vision sensor is 5W, and the power consumption of the actuator motor drive is 10W, etc.

[0037] In step S102 of some embodiments, the energy consumption score is an indicator used to measure the current energy consumption status of the robot, usually expressed in the range of 0 to 100. In this embodiment, the lower the energy consumption score, the greater the energy consumption pressure.

[0038] In this embodiment, the energy consumption score is obtained by predicting battery power and energy load data based on a deep learning model or decision tree model. For example, when the robot has low remaining power and is performing a high-load task, the energy consumption score may be low (e.g., 30 points), indicating that the robot is under high energy consumption pressure. Conversely, when the robot has sufficient remaining power and is performing a low-load task, the energy consumption score may be high (e.g., 80 points), indicating a healthier energy consumption state.

[0039] In step S103 of some embodiments, the target unit refers to a module in the robot whose performance or energy consumption load reaches a certain limit or threshold, causing a significant impact on the robot's energy consumption. For the process of determining the target unit, please refer to [link to relevant documentation]. Figure 2 In some embodiments, step S103 may include, but is not limited to, steps S201 to S203: Step S201: Obtain the energy consumption parameter threshold matching each hardware unit and obtain the battery power threshold.

[0040] In step S202, if the battery power data is less than the battery power threshold, then the battery is determined to be the target unit.

[0041] Step S203: For each hardware unit, if the energy consumption load data is greater than or equal to the energy consumption parameter threshold, then the hardware unit is determined as the target unit.

[0042] In step S201 of some embodiments, the energy consumption parameter threshold refers to the maximum energy consumption level that each hardware unit can withstand under normal and safe operation. Exceeding this threshold may cause performance degradation in the hardware unit due to overload. The battery power threshold refers to the threshold at which the battery power drops to a preset value, indicating that the battery is about to be unable to continue supporting task execution. This threshold is typically set based on the robot's battery capacity and the task requirements, and may be 20%.

[0043] In step S202 of some embodiments, if the battery power data is less than the battery power threshold, it indicates that the battery is currently in a low power state, and the battery will be identified as the target unit. If the battery power data is greater than or equal to the battery power threshold, the battery will not be identified as the target unit.

[0044] In step S203 of some embodiments, if the energy consumption load data of a certain hardware unit is greater than or equal to its preset energy consumption parameter threshold, then the hardware unit will be identified as a target unit. For example, assuming the processor's energy consumption parameter threshold is 50W, if the current processor's energy consumption load data is 60W, then the processor will be identified as a target unit. Similarly, if the sensor's energy consumption load data is 12W and the sensor's energy consumption parameter threshold is 10W, then the sensor will also be identified as a target unit.

[0045] Steps S201 to S203 as shown in the embodiments of this application automatically identify and determine the target unit based on the real-time battery power and the energy consumption load of the hardware unit, thereby performing targeted optimization scheduling.

[0046] In step S104 of some embodiments, task priority refers to the execution priority order assigned according to the importance and urgency of the task. In the embodiments of this application, task priorities are preset and stored in memory. Specifically, task priorities include high priority P1, medium priority P2, and low priority P3. At the beginning of robot design, preset priority assignments are made for each type of robot execution task, with core tasks having higher priority and secondary tasks having lower priority. For example, the task priority of path planning and obstacle avoidance tasks is P1, the task priority of object recognition tasks is P2, and the priority of secondary scene data training tasks is P3.

[0047] The target settings parameters are hardware resource scheduling parameters, including GPU computing power, sensor acquisition frequency, and calculation accuracy.

[0048] Please see Figure 3 In some embodiments, step S104 may include, but is not limited to, steps S301 to S304: Step S301: Obtain the robot's working stage type.

[0049] Step S302: If the work phase type is model training phase, the energy consumption score and task priority are determined as reference data.

[0050] Step S303: If the working stage type is the application stage, the task priority and the hardware type of the target unit are used as reference data.

[0051] Step S304: Generate parameters based on reference data to obtain the target setting parameters for each hardware unit.

[0052] In step S301 of some embodiments, the working phase type includes an application phase and a model training phase. The application phase refers to the phase where the robot performs tasks in a real working environment, such as object grasping and path planning. The model training phase refers to the phase where the robot learns and optimizes algorithms, such as training deep learning models and optimizing behavioral decisions.

[0053] In step S302 of some embodiments, the reference data refers to the task scheduling basis obtained by evaluating the current state of the robot at a specific work stage.

[0054] In step S303 of some embodiments, when the working stage is the application stage, the reference data refers to parameters adjusted based on the urgency of the robot task and the allocation of hardware resources. In this case, the reference data includes task priority and the hardware type of the target unit.

[0055] In step S304 of some embodiments, when the reference data includes task priority and the hardware type of the target unit (i.e., the working stage type is application stage), please refer to... Figure 4 Step S304 may include, but is not limited to, steps S401 to S403: Step S401: Query the hardware type of the target unit according to the preset hardware type parameter mapping relationship to obtain the target priority and the reference setting parameters matching the target priority.

[0056] Step S402: Select the execution task type with the task priority of the target priority as the target task type, and determine the reference unit from the hardware unit based on the target task type.

[0057] Step S403: Determine the target setting parameters based on the reference setting parameters of the reference cell and the original setting parameters of the original cell.

[0058] In step S401 of some embodiments, the hardware type parameter mapping relationship refers to a pre-established mapping rule between hardware type and task priority based on the type, function, and power consumption characteristics of each hardware unit. For example, a processor is assigned a higher priority because it plays a crucial role in the computation process. Conversely, a sensor has a lower priority if its task is auxiliary (e.g., multiple redundantly designed sensors of the same type). The target priority refers to the priority corresponding to the hardware type.

[0059] Reference setting parameters refer to the setting parameters determined for the target hardware unit based on the type of the target hardware unit and the task priority, under a preset hardware type parameter mapping relationship. These parameters include, but are not limited to, power, operating frequency, and other configurations. For example, if the high-power module is a GPU, for medium-priority tasks in P2 (such as ordinary object recognition), "model quantization technology" (converting 32-bit floating-point precision to 16-bit) is used to reduce the GPU's computing power requirements. For core tasks in P1 (such as real-time obstacle avoidance inference), only key feature channels are retained (unnecessary texture recognition channels are closed), while the inference frame rate is reduced (from 30fps to 20fps to ensure that decision accuracy is not affected).

[0060] In step S402 of some embodiments, the target task type is a task type whose task priority is consistent with the target priority. The reference unit refers to the hardware unit required to execute a task of the target task type.

[0061] In step S403 of some embodiments, the original unit is a hardware unit within the robot that is not a reference unit.

[0062] Steps S401 to S403, as illustrated in this embodiment, achieve efficient scheduling of robot hardware resources through a precise mapping relationship between hardware type and task priority. The method of this embodiment effectively solves the problem of excessive hardware resource consumption when the robot performs different tasks, ensuring reasonable hardware resource scheduling between high-priority and low-priority tasks.

[0063] When the reference data includes energy consumption scores and task priorities (i.e., the work phase type is model training phase), please refer to [link / reference]. Figure 5 In some embodiments, step S304 may also include, but is not limited to, steps S501 to S502: Step S501: Determine the parameter task mapping relationship based on the energy consumption score and the score threshold.

[0064] Step S502: Determine the target priority from the task priorities based on the energy consumption score and the preset score threshold, and map the target priority according to the parameter task mapping relationship to obtain the target setting parameters.

[0065] In step S501 of some embodiments, the parameter task mapping relationship refers to establishing a mapping relationship between task type and hardware resource configuration based on the robot's energy consumption score and a preset score threshold. The score threshold can be 60, and this embodiment does not strictly limit its specific data. For example, if the energy consumption score is <60 (energy shortage), the corresponding parameter task mapping relationship is determined. If the energy consumption score is ≥60 (energy sufficient), another parameter task mapping relationship is determined.

[0066] In step S502 of some embodiments, if the energy consumption score is less than a preset score threshold, the target priority includes high priority, medium priority, and low priority. For example, if the energy consumption score is <60 (energy shortage), the batch size is reduced (from 256 to 128) and the number of training rounds is reduced (from 100 rounds to 50 rounds) for low-priority tasks of P3 (such as training secondary scenario data); for core tasks of P1 (such as training a safety decision model), a "parameter freezing technique" is adopted (freezing non-core layer parameters and training only the output layer), while the learning rate is reduced by 10% to reduce the amount of iteration computation.

[0067] If the energy consumption score is greater than or equal to a preset score threshold, the target priority includes lower priorities. For example, if the energy consumption score is ≥80 (sufficient energy), the core parameters remain unchanged, and only the training efficiency of the P3 task is optimized.

[0068] The target priority is mapped according to the parameter task mapping relationship to obtain the target setting parameters. Taking the example of an embodiment where the energy consumption score is less than the preset score threshold, the target setting parameters can be a training batch size of 128 and a training round of 50, etc.

[0069] Steps S501 to S502 shown in the embodiments of this application can ensure that critical tasks are executed first when resources are scarce, avoid low-priority tasks from excessively consuming hardware resources, thereby effectively improving the robot's energy utilization and battery life, optimizing task execution efficiency, and thus achieving efficient operation of the robot.

[0070] Steps S301 to S304 of this embodiment illustrate a technical solution for dynamically adjusting hardware resource allocation based on the robot's work stage type and task requirements. This method, through dynamic adjustment of hardware resources and energy efficiency management, optimizes energy consumption control at different work stages, improves the robot's energy utilization and work efficiency, and ultimately enhances the robot's endurance and task performance stability.

[0071] In step S105 of some embodiments, hardware resource scheduling of the hardware unit based on the target setting parameters refers to dynamically adjusting the operating state of each hardware module of the robot according to the hardware resource configuration predicted in the aforementioned steps, so as to maximize energy efficiency. For example, when the task priority is P1 and the energy consumption score is high, the robot may reduce the processor frequency to a level suitable for the current task load, while reducing unnecessary sensor data acquisition frequencies (such as adjusting the acquisition frequency of the vision sensor from 30fps to 20fps) to achieve energy saving.

[0072] Please see Figure 6 Following step S105 in some embodiments, the hardware resource scheduling method of this application includes, but is not limited to, steps S601 to S604: Step S601: Update the energy load data to obtain intermediate energy consumption data.

[0073] Step S602: For each hardware unit, calculate the energy consumption reduction rate based on the energy load data and intermediate energy consumption data.

[0074] Step S603: Obtain the performance data of the hardware unit, verify the energy consumption reduction rate and performance data according to the preset scheduling iteration termination conditions, and obtain the verification results.

[0075] Step S604: If the verification result shows that at least one of the energy consumption reduction rate and performance data does not meet the scheduling iteration termination condition, the step of predicting the setting parameters based on the task priority, energy consumption score and target unit to obtain the target setting parameters, and scheduling hardware resources for the hardware unit based on the target setting parameters is repeated until both the energy consumption reduction rate and performance data meet the scheduling iteration termination condition.

[0076] In step S601 of some embodiments, intermediate energy consumption data refers to the energy consumption load data based on the current hardware unit after hardware resource scheduling is performed according to the target setting parameters.

[0077] In step S602 of some embodiments, the energy consumption reduction rate refers to the percentage reduction in energy consumption calculated by analyzing the energy consumption load data and intermediate energy consumption data of the hardware unit during the hardware resource scheduling process. For example, if the processor's energy consumption load data is 50W and the intermediate energy consumption data is 40W, then the energy consumption reduction rate is calculated to be 20%.

[0078] In step S603 of some embodiments, performance data may include metrics such as computational accuracy, inference accuracy, and GPU frequency used to measure processing power. The scheduling iteration termination condition refers to a set of preset conditions during hardware resource scheduling to determine whether resource optimization has been completed and the expected energy efficiency target has been achieved. The verification result refers to verifying whether the preset scheduling iteration termination conditions are met after each scheduling iteration, based on the energy consumption reduction rate and performance data of the hardware unit. The calculated energy consumption reduction rate and performance data are compared with the preset conditions to determine whether the target has been achieved. If both the energy consumption reduction rate and task performance meet the conditions, "verification passed" is output, and the iteration process ends; otherwise, hardware resource settings continue to be adjusted. For example, if GPU power consumption decreases from 80W to 50W, a reduction of 37.5% (greater than the preset energy consumption reduction rate threshold of 10%), and the core task performance meets the target (e.g., "P1 obstacle avoidance inference accuracy remains above 95%), then the current verification result is determined to be that both the energy consumption reduction rate and performance data meet the scheduling iteration termination conditions. If the energy consumption reduction rate is less than 10% or the core task accuracy drops below 95%, the current verification result is determined to be that the energy consumption reduction rate and performance data do not meet the scheduling iteration termination condition, until both the energy consumption reduction rate and performance data meet the scheduling iteration termination condition.

[0079] In step S604 of some embodiments, if the verification result shows that neither the energy consumption reduction rate nor the performance data meets the scheduling iteration termination condition, the process returns to steps S104 to S105.

[0080] Steps S601 to S604, as illustrated in this embodiment, accurately reflect the energy efficiency performance of each hardware unit during task execution by updating energy load data in real time and calculating the energy reduction rate. By acquiring the performance data of the hardware units and verifying it in conjunction with preset scheduling iteration termination conditions, it is possible to effectively determine whether the current scheduling meets task requirements and energy efficiency targets. If the energy reduction rate or performance data fails to meet expectations, the scheduling strategy will be re-executed to adjust the resource allocation of the hardware units, ensuring a balance between energy consumption and performance.

[0081] Please see Figure 7 This application also provides a hardware resource scheduling device for a robot, which can implement the above-described hardware resource scheduling method for a robot. The device includes: The acquisition module 701 is used to acquire the robot's battery power data and the type of task to be performed, as well as the energy consumption load data of all hardware units. The energy consumption assessment module 702 is used to assess the energy consumption level based on battery power data and energy load data, and obtain an energy consumption score. The unit determination module 703 is used to determine the target unit from multiple hardware units based on battery power data and energy consumption load data; The parameter prediction module 704 is used to obtain the task priority of the task type and predict the setting parameters based on at least one of the task priority, energy consumption score and target unit to obtain the target setting parameters. The hardware resource scheduling module 705 is used to schedule hardware resources for hardware units based on target setting parameters.

[0082] The specific implementation of the hardware resource scheduling device of this robot is basically the same as the specific implementation of the hardware resource scheduling method of the robot described above, and will not be repeated here.

[0083] This application also provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the hardware resource scheduling method for the robot described above. This electronic device can be any smart terminal, including tablet computers, in-vehicle computers, etc.

[0084] Please see Figure 8 , Figure 8 The hardware structure of an electronic device according to another embodiment is illustrated. The electronic device includes: The processor 801 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this application. The memory 802 can be implemented as a read-only memory (ROM), static storage device, dynamic storage device, or random access memory (RAM). The memory 802 can store the operating system and other applications. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 802, and the processor 801 calls and executes the hardware resource scheduling method of the robot in the embodiments of this application. The 803 input / output interface is used to implement information input and output. The communication interface 804 is used to enable communication and interaction between this device and other devices. Communication can be achieved through wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.). Bus 805 transmits information between various components of the device (e.g., processor 801, memory 802, input / output interface 803, and communication interface 804); The processor 801, memory 802, input / output interface 803, and communication interface 804 are connected to each other within the device via bus 805.

[0085] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the hardware resource scheduling method for the robot described above.

[0086] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0087] The robot hardware resource scheduling method, device, electronic device, and storage medium provided in this application collect data on the robot's battery level, task type, and hardware unit energy consumption load. Then, they assess energy consumption based on the battery level and energy consumption load data to obtain an energy consumption score. Next, they determine target units based on the battery level and energy consumption load data. Then, they determine target setting parameters based on at least one of the task priority of the task type, the energy consumption score, and the target unit. Finally, they perform hardware resource scheduling based on the target setting parameters. The method in this application effectively avoids situations where the robot cannot complete high-energy-consuming tasks due to insufficient power, and avoids unnecessary energy waste caused by excessive consumption of computing resources during low-load tasks. Ultimately, it improves the robot's energy utilization rate, not only extending its runtime but also increasing its operating efficiency, thereby reducing energy waste and enabling the robot to provide more durable and reliable services.

[0088] The embodiments described in this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided by the embodiments of this application. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of this application are also applicable to similar technical problems.

[0089] Those skilled in the art will understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of this application, and may include more or fewer steps than shown, or combine certain steps, or different steps.

[0090] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0091] Those skilled in the art will understand that all or some of the steps in the methods disclosed above, as well as the functional modules / units in the systems and devices, can be implemented as software, firmware, hardware, or suitable combinations thereof.

[0092] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0093] It should be understood that in this application, "at least one (item)" means one or more, and "more than" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0094] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of the units described above is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0095] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0096] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0097] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes multiple instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing programs, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0098] The preferred embodiments of the present application have been described above with reference to the accompanying drawings, but this does not limit the scope of the claims of the present application. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and substance of the embodiments of the present application shall be within the scope of the claims of the present application.

Claims

1. A method for scheduling hardware resources of a robot, characterized in that, The robot is equipped with at least two hardware units, and the method includes: Obtain the robot's battery power data and task type, and obtain the energy consumption load data of all hardware units; Based on the battery power data and the energy load data, the energy consumption level is evaluated to obtain an energy consumption score; The target unit is determined from the plurality of hardware units based on the battery power data and the energy load data; Obtain the task priority of the task type to be executed, and predict the setting parameters based on the task priority, the energy consumption score and at least one of the target units to obtain the target setting parameters; Hardware resource scheduling is performed on the hardware unit based on the target setting parameters.

2. The method according to claim 1, characterized in that, The step of predicting setting parameters based on at least one of the task priority, the energy consumption score, and the target unit to obtain target setting parameters includes: Obtain the working stage type of the robot; wherein, the working stage type includes the application stage and the model training stage; If the work phase type is the model training phase, the energy consumption score and the task priority will be used as reference data. If the work stage type is the application stage, the task priority and the hardware type of the target unit shall be used as reference data. Parameters are generated based on the reference data to obtain the target setting parameters for each hardware unit.

3. The method according to claim 2, characterized in that, When the reference data includes the task priority and the hardware type of the target unit, the step of generating parameters based on the reference data to obtain the target setting parameters for each hardware unit includes: The hardware type of the target unit is queried according to the preset hardware type parameter mapping relationship to obtain the target priority and the reference setting parameters matching the target priority. The execution task type with the task priority equal to the target priority is taken as the target task type, and a reference unit is determined from the hardware unit based on the target task type; The target setting parameters are determined based on the reference setting parameters of the reference unit and the original setting parameters of the original unit; wherein, the original unit is a hardware unit within the robot that is not the reference unit.

4. The method according to claim 2, characterized in that, When the reference data includes the energy consumption score and the task priority, the step of generating parameters based on the reference data to obtain the target setting parameters for each hardware unit includes: The parameter task mapping relationship is determined based on the energy consumption score and the score threshold. The target priority is determined from the task priorities based on the energy consumption score and the preset score threshold, and the target priority is mapped according to the parameter task mapping relationship to obtain the target setting parameters.

5. The method according to claim 4, characterized in that, The task priorities include high priority, medium priority, and low priority. The step of determining a target priority from the task priorities based on the energy consumption score and a preset score threshold, and mapping the target priority according to the parameter task mapping relationship to obtain the target setting parameters, includes: If the energy consumption score is less than the preset score threshold, the target priority includes high priority, medium priority and low priority; If the energy consumption score is greater than or equal to the preset score threshold, then the target priority includes low priority; The target priority is mapped according to the parameter task mapping relationship to obtain the target setting parameters.

6. The method according to any one of claims 1 to 5, characterized in that, The hardware unit includes a battery, and the step of determining the target unit from a plurality of hardware units based on the battery power data and the energy consumption load data includes: Obtain the energy consumption parameter threshold matching each of the aforementioned hardware units, and obtain the battery power threshold. If the battery power data is less than the battery power threshold, then the battery is determined to be the target unit; For each of the aforementioned hardware units, if the energy consumption load data is greater than or equal to the energy consumption parameter threshold, then the hardware unit is determined to be the target unit.

7. The method according to any one of claims 1 to 5, characterized in that, After performing hardware resource scheduling on the hardware unit based on the target setting parameters, the method further includes: The energy load data is updated to obtain intermediate energy consumption data; For each of the aforementioned hardware units, the energy consumption reduction rate is calculated based on the energy consumption load data and the intermediate energy consumption data; The performance data of the hardware unit is obtained, and the energy consumption reduction rate and the performance data are verified according to the preset scheduling iteration termination condition to obtain the verification result. If the verification result reflects that at least one of the energy consumption reduction rate and the performance data does not meet the scheduling iteration termination condition, the steps of predicting setting parameters based on at least one of the task priority, the energy consumption score and the target unit to obtain target setting parameters, and scheduling hardware resources for the hardware unit based on the target setting parameters are re-executed until both the energy consumption reduction rate and the performance data meet the scheduling iteration termination condition.

8. A hardware resource scheduling device for a robot, characterized in that, The device includes: The acquisition module is used to acquire the robot's battery power data and the type of task to be performed, and to acquire the energy consumption load data of all the hardware units. The energy consumption assessment module is used to assess the energy consumption level based on the battery power data and the energy load data, and obtain an energy consumption score. A unit determination module is used to determine a target unit from a plurality of hardware units based on the battery power data and the energy consumption load data; The parameter prediction module is configured to obtain the task priority of the task type to be executed, and to predict the setting parameters based on at least one of the task priority, the energy consumption score and the target unit to obtain the target setting parameters. The hardware resource scheduling module is used to schedule hardware resources for the hardware unit based on the target setting parameters.

9. An electronic device, characterized in that, The electronic device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the method according to any one of claims 1 to 7.

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

Images