A power management method, device and humanoid robot

Abstract

The application discloses a power management method, comprising the following steps: S1. receiving collected detection data; S2. comparing current data and voltage data with a normal working threshold range, and marking data out of the range; S3. generating an abnormal signal when it is detected that the current data is greater than the normal working threshold range for a continuous preset number of sampling periods, or it is detected that the voltage data is lower than the normal working threshold range for a continuous preset number of sampling periods; S4. receiving abnormal detection noise data collected by an abnormal noise detection signal, performing feature recognition on the abnormal detection noise data, generating a judgment result, and sending the judgment result to a main controller, so that the main controller generates a corresponding power control signal after receiving the judgment result. The application also discloses a power management device and a humanoid robot. Compared with the prior art, the application realizes efficient, accurate and anti-interference energy management.

Description

Technical Field

[0001] This invention relates to the field of battery power technology, and in particular to a power management method, device and humanoid robot. Background Technology

[0002] With the development of humanoid robot technology, their functions are becoming increasingly complex, and the number of integrated electrical components (such as joint motors, sensors, control systems, and actuators) has increased significantly. Furthermore, the power requirements of each component vary greatly, and their operating states change dynamically (e.g., sudden changes in leg joint load during walking, and fluctuations in arm motor power during grasping). This places extremely high demands on battery energy management: it requires real-time monitoring of the energy consumption status of each component, precise energy allocation, and the ability to cope with complex electromagnetic environments and mechanical vibration interference.

[0003] Existing battery energy management technologies are relatively mature in two-wheeled vehicles or new energy vehicles, but humanoid robots differ significantly from these devices:

[0004] Load complexity: The electrical components of humanoid robots are distributed in a dispersed manner (all joints, end effectors, etc.), and the load changes dynamically with the movement, making the response speed of traditional centralized monitoring insufficient;

[0005] Interference problem: Electrical appliances (especially motors) generate mechanical vibration and electromagnetic noise when they are working, which directly affects the detection accuracy of sensors and leads to distortion of energy consumption data;

[0006] Sensor protection: During the movement of a humanoid robot, external environmental factors (such as collisions and dust) may damage exposed sensors, affecting long-term stability;

[0007] Current and voltage feedback lag: Current and voltage detection of multi-load electrical appliances in the present technology mostly adopts periodic sampling, which makes it difficult to capture instantaneous power changes (such as the current peak when the joint stops suddenly), which can easily lead to delays in energy management commands.

[0008] Therefore, there is an urgent need for a power management method adapted to the characteristics of humanoid robots to achieve efficient, accurate, and interference-resistant energy management, especially to enhance the real-time performance of current and voltage detection and the dynamic response capability of feedback adjustment. Summary of the Invention

[0009] The purpose of this invention is to provide a power management method, device, and humanoid robot to achieve efficient, accurate, and interference-resistant energy management, especially to enhance the real-time performance of current and voltage detection and the dynamic response capability of feedback adjustment.

[0010] The above-mentioned technical objective of the present invention is achieved through the following technical solution:

[0011] A power management method includes the following steps:

[0012] S1. Receive the collected detection data, including current data and voltage data;

[0013] S2. Compare the current and voltage data with the normal operating threshold range and mark the data that exceeds the range;

[0014] S3. When the current data is detected to be greater than the normal operating threshold range for a consecutive preset number of sampling cycles, or when the voltage data is detected to be lower than the normal operating threshold range for a consecutive preset number of sampling cycles, an abnormal signal is generated and sent to the main controller, and an abnormal noise detection signal is issued at the same time.

[0015] S4. Receive abnormal noise data triggered by the abnormal noise detection signal, perform feature recognition on the abnormal noise data, generate a judgment result based on the feature recognition result and combined with the real-time collected current and voltage data, and send the judgment result to the main controller, so that the main controller generates the corresponding power control signal after receiving the judgment result.

[0016] In a preferred embodiment, in step S4, the feature recognition result includes normal noise signal and high-frequency noise signal. When the current data is greater than 150% of the maximum value of the normal operating threshold range and is a high-frequency noise signal, the judgment result is mechanical jamming. When the voltage data is less than 80% of the minimum value of the normal operating threshold range and is a normal noise signal, the judgment result is a decrease in battery power supply capacity. When the power fluctuation is greater than ±30% within a preset time and is a normal noise signal, the judgment result is poor circuit contact.

[0017] In a preferred embodiment, in each step, when current data and voltage data are received, the current data and voltage data are filtered.

[0018] In a preferred embodiment, in step S1, the normal sampling frequency is 1kHz. When the current change rate exceeds 5A / ms or the voltage fluctuation exceeds ±5%, the sampling frequency is increased to 10kHz, and sampling is continued for 50ms before returning to the normal frequency.

[0019] In a preferred embodiment, step S4 further includes the following steps before feature recognition: identifying basic sounds; if there are basic sounds that match the features, they are considered valid noise data, and the basic sounds are stripped before feature recognition is performed; if there are no basic sounds that match the features, they are considered invalid noise data, and feature recognition is not performed.

[0020] The present invention also provides a power management device, comprising:

[0021] A current sensor is used to collect current data from electrical appliances.

[0022] Voltage sensors are used to collect voltage data from electrical appliances;

[0023] Noise sensors are used to collect noise data from electrical appliances;

[0024] The detection unit is used to receive data information collected by the current sensor, voltage sensor and noise sensor, and to execute the power management method described above.

[0025] In a preferred embodiment, the current sensor is a Hall effect sensor and is installed in the power supply circuit of the appliance; the voltage sensor is installed in the power supply interface of the appliance; and the noise sensor is located at the position of the moving part of the appliance.

[0026] In a preferred embodiment, the noise sensor is mounted on a floating platform. The floating platform includes a housing and a lift fan. The housing includes a floating cover and a base. The base has a motor chamber and a fan chamber. The side of the motor chamber has an air inlet with a filter screen. The fan chamber is connected to the motor chamber and has a guide hole with an annular limiting platform in the guide hole. The floating cover has a guide rod inserted into the guide hole, and the end of the guide rod has a limiting block. The noise sensor is mounted on the bottom of the floating cover. The lift fan includes a drive motor and fan blades connected to each other. The drive motor is located in the motor chamber, and the fan blades are located in the fan chamber. The operating sound of the lift fan is a basic tone.

[0027] In a preferred embodiment, a sound-silencing pad is provided on the annular limiting platform, a magnet is provided at the bottom of the guide hole, and the limiting block is made of magnetic material.

[0028] The present invention also provides a humanoid robot that applies the power management method described above, or includes the power management device described above.

[0029] Compared with existing technologies, this invention combines current, voltage, and noise data for anomaly detection, effectively improving the accuracy and specificity of identifying abnormal electrical appliance states. When relying solely on current and voltage data, anomalies can be caused by various factors. For example, a current exceeding a normal threshold could be a normal instantaneous load fluctuation, or it could be due to mechanical jamming, component wear, or other faults, making accurate differentiation difficult using only electrical parameters. Conversely, a voltage below a threshold could be a temporary fluctuation in the power supply circuit, or a serious problem such as an internal short circuit. Introducing noise data allows for the capture of acoustic features corresponding to anomalies—for instance, mechanical jamming generates high-frequency friction noise, component wear produces periodic abnormal noises, normal load fluctuations typically do not involve abnormal noise, and short circuits may generate specific arc noise. This association of abnormal electrical parameters with specific fault types avoids misjudgments caused by a single abnormal electrical parameter, reduces unnecessary power adjustment actions by the main controller, and improves system stability.

[0030] Meanwhile, this multi-dimensional data fusion-based judgment method can refine the anomaly types, making the power control signals generated by the main controller more targeted. For example, when the current is abnormal and accompanied by mechanical jamming noise, the main controller can directly trigger the shutdown protection to prevent further damage to the components; if the current is abnormal but the noise does not change significantly, it may be judged as a poor contact in the power supply circuit. In this case, the main controller can adjust the voltage output and issue a maintenance signal instead of directly shutting down, which not only ensures equipment safety but also reduces functional interruptions caused by misjudgment, improving the operating efficiency and fault tolerance of the humanoid robot's power system.

[0031] Furthermore, the mechanism of triggering noise detection by "continuously preset sampling periods" can avoid frequent noise detection and reduce system resource consumption. Once triggered, combined with the rapid identification of noise characteristics, it can ensure timely response while maintaining judgment accuracy, so that power management can achieve a balance between efficiency and accuracy. This is especially suitable for complex working conditions of humanoid robots with multiple loads and dynamic changes.

[0032] The aforementioned processes, such as data reception, comparison marking, abnormal signal generation, noise data processing, and feature recognition, are all performed by independent detection units without the need for the main controller to participate in intermediate calculations. This design can significantly reduce the computational load of the main controller, preventing it from consuming too many resources due to processing large amounts of raw data. This allows the main controller to focus more on the generation and execution of core control commands, further enhancing the real-time response capability of the entire power management system. Attached Figure Description

[0033] Figure 1 This invention relates to a flowchart of a power management method.

[0034] Figure 2This invention relates to a schematic diagram of the system structure used in a power management method.

[0035] Figure 3 This is a schematic diagram of the structure of a battery used in a power management method according to the present invention.

[0036] Figure 4 This is a schematic diagram of the structure of a floating platform for a power management device, which relates to the present invention.

[0037] Figure 5 This is a schematic diagram of the structure of a floating platform of a power management device after the floating cover is raised, according to the present invention.

[0038] Figure 6 This invention relates to a schematic diagram of the internal structure of a floating platform for a power management device.

[0039] Battery 1; Energy Management Unit 2; Main Controller 3; Detection Unit 4; Current Sensor 5; Voltage Sensor 6; Noise Sensor 7; Electrical Appliance 8; Floating Cover 9; Guide Rod 10; Limiting Block 11; Base 12; Motor Chamber 13; Air Inlet 14; Filter 15; Fan Chamber 16; Guide Hole 17; Annular Limiting Platform 18; Magnet 19; Silent Pad 20; Drive Motor 21; Fan Blade 22. Detailed Implementation

[0040] The present invention will be further described in detail below with reference to the accompanying drawings.

[0041] This specific embodiment is merely an explanation of the present invention and is not intended to limit the invention. After reading this specification, those skilled in the art can make modifications to this embodiment without contributing any inventive step, but such modifications are protected by patent law as long as they are within the scope of the claims of the present invention. Example 1

[0042] like Figures 1 to 3 As shown, a power management method includes the following steps:

[0043] S1. Receive the collected detection data, including current data and voltage data;

[0044] S2. Compare the current and voltage data with the normal operating threshold range and mark the data that exceeds the range;

[0045] S3. When the current data is detected to be greater than the normal operating threshold range for a consecutive preset number of sampling cycles, or when the voltage data is detected to be lower than the normal operating threshold range for a consecutive preset number of sampling cycles, an abnormal signal is generated and sent to the main controller 3, and an abnormal noise detection signal is issued at the same time.

[0046] S4. Receive abnormal noise detection data triggered by the abnormal noise detection signal, perform feature recognition on the abnormal noise detection data, generate a judgment result based on the feature recognition result and combined with the real-time collected current and voltage data, and send the judgment result to the main controller 3, so that the main controller 3 generates the corresponding power control signal after receiving the judgment result.

[0047] Combining current, voltage, and noise data to identify anomalies significantly improves the accuracy and discriminative power of identifying abnormal states in appliance 8. Relying solely on current and voltage data, anomalies can stem from various causes—for example, current exceeding the normal range could be due to normal instantaneous load fluctuations in appliance 8, or faults such as mechanical jamming or component wear; voltage below the threshold could be due to temporary fluctuations in the power supply circuit, or serious problems such as internal short circuits in appliance 8. These differences are difficult to accurately distinguish using only electrical parameters. However, by introducing noise data, acoustic characteristics corresponding to anomalies can be captured through feature recognition: mechanical jamming produces high-frequency friction noise, component wear produces periodic abnormal noises, normal load fluctuations typically have no abnormal noise, and short circuits may be accompanied by specific arc noises. This allows for the association of abnormal electrical parameters with specific fault types, avoiding misjudgments caused by a single abnormal electrical parameter, reducing unnecessary power adjustments by the main controller 3, and enhancing system stability.

[0048] Meanwhile, this multi-dimensional data fusion-based judgment method can refine the specific types of anomalies, allowing the power control signals generated by the main controller 3 to be more accurately adapted to the actual situation. For example, when the current is abnormal and accompanied by mechanical jamming noise, the main controller 3 can directly trigger the shutdown protection to prevent component damage; if the current is abnormal but the noise does not change significantly, it may be due to poor contact in the power supply circuit. In this case, the main controller 3 can adjust the voltage output and issue a maintenance signal instead of directly shutting down, which ensures equipment safety, reduces functional interruptions caused by misjudgment, and improves the operating efficiency and fault tolerance of the humanoid robot's power system.

[0049] Furthermore, the mechanism of triggering noise detection by "continuously preset sampling periods" can avoid frequent noise detection and reduce system resource consumption. After triggering, the rapid identification of noise characteristics can ensure both judgment accuracy and response speed, thus achieving a balance between efficiency and accuracy in power management. This is especially suitable for complex working scenarios of humanoid robots with multiple loads and dynamic changes.

[0050] The processes of data reception, comparison marking, abnormal signal generation, noise data processing, and feature recognition are all completed by the independent detection unit 4, and the main controller 3 does not participate in these intermediate calculation steps. This design can significantly reduce the computational pressure on the main controller 3, avoid it from consuming too many resources due to processing massive amounts of raw data, and allow the main controller 3 to focus more on the generation and execution of core control commands, thereby further improving the real-time response performance of the entire power management system.

[0051] In step S4, the feature recognition results include normal noise signals and high-frequency noise signals. When the current data is greater than 150% of the maximum value of the normal operating threshold range and is a high-frequency noise signal, the judgment result is mechanical jamming. When the voltage data is less than 80% of the minimum value of the normal operating threshold range and is a normal noise signal, the judgment result is a decrease in the power supply capacity of battery 1. When the power fluctuation is greater than ±30% within a preset time and is a normal noise signal, the judgment result is poor circuit contact.

[0052] By specifically correlating feature recognition results with current, voltage, and power parameters, precise judgment rules for specific fault types are formed, enabling refined differentiation and localization of abnormal states. When the current data exceeds the normal threshold by 150% and is accompanied by high-frequency noise, it is determined to be mechanical jamming, which can directly lock the abnormality of the mechanical transmission system and provide a clear basis for the main controller 3 to trigger immediate shutdown protection, avoiding component wear or damage caused by the continued jamming state; when the voltage is lower than the normal threshold by 80% and the noise is normal, it is determined to be a decrease in the power supply capacity of battery 1, which can accurately distinguish between the performance degradation of the power supply system itself and the fault of appliance 8, allowing the main controller 3 to prioritize adjusting the energy distribution strategy (such as reducing the power of non-core loads) to ensure the continuous operation of critical functions; when the power fluctuation exceeds ±30% and the noise is normal, it is determined to be poor circuit contact, which can accurately identify the connection problem of the power supply circuit, providing a basis for the main controller 3 to issue maintenance prompts rather than emergency shutdown, reducing functional interruptions caused by misjudgment.

[0053] This specific fault classification method based on multi-parameter correlation avoids general judgments of abnormal states, making the power control signal of the main controller 3 more consistent with the actual fault scenario. This not only improves the pertinence and effectiveness of fault handling, but also reduces unnecessary protective actions from interfering with the normal operation of the robot, further optimizing the reliability and operating efficiency of the power management system. It is especially suitable for the needs of humanoid robots for rapid fault location and accurate response under complex load environments.

[0054] In each step, upon receiving current and voltage data, the data is filtered. Filtering effectively eliminates or reduces interference components in the data, including electromagnetic noise generated by the appliance 8 during operation, measurement noise from the sensor itself, and high-frequency noise introduced by circuit conduction, making the collected current and voltage data closer to the true values. This processing avoids false anomaly signals caused by noise—for example, unfiltered current data may exhibit spikes due to transient electromagnetic interference, falsely triggering a "current exceeds threshold" judgment; voltage data containing high-frequency fluctuations may be misjudged as unstable power supply. After filtering, the smoothness and reliability of the data are significantly improved, making subsequent comparisons with normal operating thresholds more accurate, anomaly judgments in continuous sampling periods more reliable, and reducing false triggers caused by data distortion.

[0055] Possible filtering methods include: Kalman filtering, suitable for dynamically changing current and voltage signals, which can suppress Gaussian noise in real time through a prediction-update mechanism, balancing real-time performance and filtering effect; moving average filtering, which can effectively smooth high-frequency noise by averaging multiple consecutive sampling points, with a simple and easy-to-implement algorithm, suitable for detection units with limited computing power; wavelet filtering, which can filter out noise while preserving the characteristics of signal abrupt changes (such as real current surges), especially suitable for scenarios that need to distinguish between real load abrupt changes and interference; and finite impulse response (FIR) filtering, which has linear phase characteristics and can accurately filter out electromagnetic interference of specific frequencies (such as fixed-frequency noise generated by motor PWM), further improving data purity.

[0056] Furthermore, in step S1, the normal sampling frequency is 1kHz. When the current change rate exceeds 5A / ms or the voltage fluctuation exceeds ±5%, the sampling frequency is increased to 10kHz, and sampling continues for 50ms before returning to the normal frequency. This design of dynamically adjusting the sampling frequency achieves a balance between efficiency and accuracy. The normal 1kHz sampling reduces data volume and computational load, avoids redundant information consuming resources, and meets the basic monitoring needs under stable operating conditions. When the current change rate exceeds 5A / ms or the voltage fluctuation exceeds ±5%, such transient changes may imply sudden load changes or potential faults. Increasing to 10kHz high-frequency sampling at this time allows for dense recording of transient characteristics within 50ms, avoiding missing critical details such as sudden current increases and voltage drops. Returning to the normal frequency after 50ms prevents excessive resource consumption from prolonged high-frequency sampling, ultimately providing more comprehensive and accurate raw data for subsequent threshold comparison and anomaly judgment, improving the system's response capability to complex operating conditions.

[0057] Furthermore, in step S4, before feature recognition, the following steps are included: identifying the basic sound. If a basic sound matching the characteristics exists, it is considered valid noise data, and the basic sound is stripped before feature recognition. If no basic sound matching the characteristics exists, it is considered invalid noise data, and feature recognition is not performed. By identifying the basic sound (such as the characteristic sound of a lift fan), it can be verified whether the floating platform is raised normally and whether the noise sensor 7 is in an effective detection position. If a basic sound matching the characteristics exists, it indicates that the noise data comes from an effective environment without mechanical vibration interference, and the reliability of the data is confirmed. Stripping the basic sound at this time can eliminate the interference of the fan's own sound, allowing subsequent feature recognition to focus more on the real noise of the electrical appliance 8 (such as the high-frequency sound of mechanical jamming or the specific noise of circuit abnormality), thus improving the accuracy of recognition. If no basic sound exists, the data is directly determined to be invalid and feature recognition is not performed. This avoids the mechanical vibration noise caused by the floating platform not being raised (such as fan failure) from being mixed into the detection results, preventing erroneous data from causing misjudgment. At the same time, it reduces the redundant processing of invalid data by the detection unit 4, saves computing resources, and ultimately provides a more reliable noise basis for multi-parameter fusion judgment. Example 2

[0058] like Figures 1 to 6 As shown, a power management device includes:

[0059] Current sensor 5 is used to collect current data from electrical appliance 8;

[0060] Voltage sensor 6 is used to collect voltage data from electrical appliance 8;

[0061] Noise sensor 7 is used to collect noise data from electrical appliance 8;

[0062] The detection unit 4 is used to receive data information collected by the current sensor 5, voltage sensor 6 and noise sensor 7, and to execute the power management method described in Example 1.

[0063] The first power management device in this embodiment uses three types of sensors—current, voltage, and noise—to collaboratively collect data, achieving multi-dimensional monitoring of the electrical appliance 8's status: current and voltage sensors 6 capture electrical parameter characteristics, and noise sensors 7 supplement acoustic information. The combination of these three sensors comprehensively reflects the operating status of the electrical appliance 8, avoiding the limitations of single-parameter detection. The detection unit 4, as the core processing module, independently executes the power management method, reducing the burden on the main controller 3 and ensuring data accuracy and processing efficiency through dynamic sampling, filtering, and basic sound recognition mechanisms. The overall device's hardware and algorithms work together to achieve closed-loop management from data acquisition to anomaly detection, improving the timeliness and reliability of power adjustment and adapting to the complex and varied load scenarios of the humanoid robot's electrical appliance 8.

[0064] The current sensor 5 is a Hall effect sensor and is installed in the power supply circuit of the appliance 8. The voltage sensor 6 is installed in the power supply interface of the appliance 8, and the noise sensor 7 is located at the moving part of the appliance 8. The Hall effect current sensor 5 has a fast response (≤1μs) and high accuracy (±0.5%). Installed in the power supply circuit, it can directly capture the real-time current changes of the appliance 8, and is especially suitable for transient current detection in dynamic scenarios such as motor start-up, reducing signal attenuation caused by line transmission. The voltage sensor 6 is installed in the power supply interface and can directly collect the actual voltage value that the appliance 8 is subjected to, avoiding measurement deviations caused by line voltage drop and ensuring that the voltage data is consistent with the actual working state of the appliance 8. The noise sensor 7 is located at the moving part and can collect the original noise generated by mechanical movement (such as joint friction and motor running sound) at close range, reducing environmental noise interference and making the correlation between noise data and the operating state of the appliance 8 stronger. The adaptive design of the installation positions and types of the three sensors improves the authenticity and correlation of current, voltage, and noise data.

[0065] The noise sensor 7 is mounted on a floating platform. The floating platform includes a housing and a lift fan. The housing includes a floating cover 9 and a base 12. The base 12 has a motor chamber 13 and a fan chamber 16. The side of the motor chamber 13 has an air inlet 14, and a filter screen 15 is provided on the air inlet 14. The fan chamber 16 is connected to the motor chamber 13. The fan chamber 16 has a guide hole 17, and an annular limiting platform 18 is provided in the guide hole 17. The floating cover 9 has a guide rod 10, which is inserted into the guide hole 17. The end of the guide rod 10 has a limiting block 11. The noise sensor 7 is mounted on the bottom of the floating cover 9. The lift fan includes a drive motor 21 and a fan blade 22 connected to each other. The drive motor 21 is located in the motor chamber 13, and the fan blade 22 is located in the fan chamber 16. The operating sound of the lift fan is a basic tone. The floating cover 9 is lifted by a lifting fan, which separates the noise sensor 7 from the base 12 and surrounding mechanical parts, physically isolating vibration transmission and eliminating the interference of mechanical vibration on noise detection in space, ensuring that the collected noise data only reflects the working status of the electrical appliance 8; the cooperation between the guide rod 10 and the guide hole 17 ensures stable lifting, and the limit block 11 and the annular limit platform 18 prevent excessive displacement and improve structural reliability.

[0066] After the lift fan is turned off, the floating cover 9 falls back to the base 12, and the noise sensor 7 is enclosed in the sealed space formed by the housing. The filter 15 blocks dust and impurities from entering. With the protection of the housing, it effectively isolates external damage such as collisions and liquids, and significantly extends the life of the sensor.

[0067] Using the sound of the lifting fan as the base tone, it can provide a self-test benchmark for the lifting and lowering status of the floating platform (the presence or absence of a base tone determines whether it is working normally), and can also identify zero-point drift (such as fixed frequency deviation) by comparing the benchmark sound wave with the signal collected by the sensor, so as to facilitate timely calibration and continuously ensure the accuracy of noise detection.

[0068] The annular limiting platform 18 is equipped with a sound-absorbing pad 20, and the bottom of the guide hole 17 is equipped with a magnet 19. The limiting block 11 is made of magnetic material. The sound-absorbing pad 20 on the annular limiting platform 18 can prevent the limiting block 11 from rigidly colliding with the annular limiting platform 18 when the floating cover 9 is raised through elastic buffering. It effectively absorbs the vibration and noise at the moment of contact, prevents the additional noise generated in this process from interfering with the noise sensor 7's collection of the working sound of the electrical appliance 8, and ensures the purity of the noise data.

[0069] The cooperation between the magnet 19 at the bottom of the guide hole 17 and the magnetic limiting block 11 achieves bidirectional optimization: when the fan is off, the magnetic force attracts the limiting block 11, making the floating cover 9 fit tightly against the base 12, preventing the floating cover 9 from shifting due to shaking caused by the robot's movement, ensuring the sealing of the enclosed space, and strengthening the protection of the noise sensor 7; when the fan is on, the lift and magnetic force cancel each other out, achieving a better levitation state.

[0070] Regarding the installation of the floating platform, it can be installed by attaching an adhesive layer to the base 12, or by setting a mounting bracket on the base 12 that can be fixed with screws. Example 3

[0071] like Figures 1 to 6 As shown, a humanoid robot applies the power management method described in Embodiment 1, or includes the power management device described in Embodiment 2, to achieve a better level of power management.

[0072] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal device. Unless otherwise specified, an element defined by the phrase "comprising..." or "including..." does not exclude the presence of additional elements in the process, method, article, or terminal device that includes said element. Additionally, in this document, "greater than," "less than," "exceeding," etc., are understood to exclude the stated number; "above," "below," "within," etc., are understood to include the stated number.

[0073] The above description of the embodiments is provided to facilitate understanding and use of the present invention by those skilled in the art. It is obvious to those skilled in the art that various modifications can be easily made to the embodiments, and the general principles described herein can be applied to other embodiments without creative effort. Therefore, the present invention is not limited to the above embodiments. Improvements and modifications made by those skilled in the art based on the disclosure of the present invention without departing from the scope of the present invention should be within the protection scope of the present invention.

Claims

1. A power management method, characterized by, Comprising the following steps: S1. Receiving collected detection data, the detection data including current data and voltage data; S2. Comparing the current data and voltage data with the normal working threshold range, and marking the data out of range; S3. When the current data is greater than the normal working threshold range for a continuous preset number of sampling periods, or the voltage data is lower than the normal working threshold range for a continuous preset number of sampling periods, an abnormal signal is generated and sent to the main controller, and an abnormal noise detection signal is also sent out; S4. Receiving abnormal detection noise data collected triggered by the abnormal noise detection signal, and performing feature recognition on the abnormal detection noise data, generating a judgment result according to the feature recognition result, combining the real-time collected current data and voltage data, and sending the judgment result to the main controller, so that the main controller generates a corresponding power control signal after receiving the judgment result; In step S4, the feature recognition result includes normal noise signal and high-frequency noise signal, when the current data is greater than 150% of the maximum value of the normal working threshold range, and is a high-frequency noise signal, the judgment result is mechanical jamming, when the voltage data is less than 80% of the minimum value of the normal working threshold range, and is a normal noise signal, the judgment result is battery power supply capability decline, when the power fluctuation is greater than ±30% within a preset time, and is a normal noise signal, the judgment result is circuit contact failure.

2. The power management method of claim 1, wherein, In each step, when receiving the current data and voltage data, the current data and voltage data are filtered.

3. The power management method of claim 1, wherein, In step S1, the normal sampling frequency is 1 kHz, when the current change rate exceeds 5A / ms or the voltage fluctuation exceeds ±5%, the sampling frequency is increased to 10 kHz, and the normal frequency is restored after continuous collection of 50ms.

4. The power management method of claim 1, wherein, In step S4, before feature recognition, it further includes the steps of identifying the basic sound, if there is a basic sound that meets the characteristics, it is effective noise data, and the basic sound is stripped, and then feature recognition is performed; If there is no basic sound that meets the characteristics, it is invalid noise data, and feature recognition is not performed.

5. A power management apparatus, characterized by comprising: Comprising: A current sensor for collecting current data of an electrical appliance; A voltage sensor for collecting voltage data of an electrical appliance; A noise sensor for collecting noise data of an electrical appliance; A detection unit for receiving data information collected by the current sensor, voltage sensor and noise sensor, and executing the power management method of any one of claims 1 to 4.

6. The power management device of claim 5, wherein, The current sensor adopts a Hall effect sensor and is installed in the power supply circuit of the electrical appliance, the voltage sensor is installed in the power supply interface of the electrical appliance, and the noise sensor is arranged at the position of the moving part of the electrical appliance.

7. The power management device of claim 5, wherein, The noise sensor is installed on a floating platform, the floating platform comprises a shell and a lift fan, the shell comprises a floating cover and a base, a motor chamber and a fan chamber are arranged in the base, a side of the motor chamber is provided with an air inlet, a filter screen is arranged on the air inlet, the fan chamber is communicated with the motor chamber, a guide hole is arranged on the fan chamber, an annular limiting table is arranged in the guide hole, a guide rod is arranged on the floating cover and inserted into the guide hole, a limiting block is arranged at the end of the guide rod, the noise sensor is installed at the bottom of the floating cover, the lift fan comprises a driving motor and a fan blade which are connected with each other, the driving motor is arranged in the motor chamber, the fan blade is arranged in the fan chamber, and the running sound of the lift fan is a basic sound.

8. The power management device of claim 7, wherein, A mute pad is arranged on the annular limiting table, a magnet is arranged at the bottom of the guide hole, and the limiting block is made of a magnetic material.

9. A humanoid robot, characterized by, A power management method according to any one of claims 1 to 4, or a power management device according to any one of claims 5 to 8.

Images