A motor control method and system for underwater pollution cleaning operations of intelligent robots

Abstract

The present invention discloses a motor control method and system for underwater pollution cleaning operations of intelligent robots, and specifically relates to the field of underwater operations, including sensor deployment, data preprocessing, model analysis, and motor control. The present invention realizes accurate evaluation of the motor state by comprehensively collecting key parameters of underwater pollution cleaning operations of intelligent robots, and performing fine data preprocessing and model analysis. At the same time, based on the evaluation results, the present invention can adjust the motor control instructions in real time and flexibly, effectively avoiding the occurrence of motor failures and ensuring the smooth progress of underwater pollution cleaning operations. In addition, the present invention also provides detailed data recording and analysis functions, which provides strong support for subsequent system maintenance and optimization, and significantly improves the efficiency and reliability of underwater pollution cleaning operations of intelligent robots.

Description

Technical Field

[0001] The present invention relates to the technical field of underwater operations, and more specifically, to a motor control method and system for underwater pollution cleaning operations of an intelligent robot. Background Art

[0002] Intelligent robot technology has made great progress in many fields in recent years, especially in the field of underwater operations, and its application prospects are becoming increasingly broad. Underwater pollution removal is an important direction for the application of intelligent robots, which is crucial to maintaining the cleanliness and ecological balance of the water environment. Due to the complex and changeable underwater environment, extremely high requirements are placed on the motor control system of intelligent robots.

[0003] At present, most intelligent robot underwater pollution cleaning operation systems use preset programs to control the operation of motors. This process usually includes the following steps: First, the system receives pollution cleaning instructions from the operator. These instructions may include parameters such as the pollution cleaning area, operation depth, and pollution cleaning speed. Then, the system calculates the speed and power that the motor should reach based on the preset algorithm and these parameters. Then, the system transmits these control signals to the motor through the motor drive circuit, thereby driving the robot to complete the pollution cleaning operation. During the operation, the system will continuously collect motor operation data and underwater environment data through sensors in order to monitor and adjust the motor status.

[0004] Although the existing technology has achieved certain results, there are still some obvious problems in practical applications. On the one hand, due to the complexity and uncertainty of the underwater environment, the existing system is often difficult to accurately predict and adapt to various changes, resulting in inaccurate and unstable motor control. For example, in areas with turbulent water flow or large changes in water depth, the motor may overheat or fail due to excessive load. On the other hand, the existing system has limitations in data processing and decision-making, and it is difficult to make full use of the collected data to optimize the motor control strategy, which leads to low operating efficiency and prone to failure. In addition, the maintenance cost of the existing system is high, and there are also high requirements for the technical level of the operator. Summary of the invention

[0005] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a motor control method and system for underwater pollution cleaning operations of an intelligent robot, and solves the problems raised in the above-mentioned background technology through the following scheme.

[0006] To achieve the above object, the present invention provides the following technical solution: a motor control method for underwater pollution cleaning operation of an intelligent robot, comprising:

[0007] Step 1: Sensor deployment: used to deploy sensors on the target robot to collect motor operation parameters, hydrological environment parameters, posture dynamics parameters and operation efficiency parameters of the target robot's underwater pollution cleaning operation, and transmit the collected parameters to step 2;

[0008] Step 2: Data preprocessing: It is used to preprocess the data transmitted in step 1, including standardization, time alignment, data quality control, environmental factor compensation and signal processing, and transmit the preprocessed data to step 3;

[0009] Step 3: Model analysis: used to establish a mathematical model to analyze the data transmitted in step 2, and transmit the analysis results to step 4;

[0010] Step 4: Motor control: Determine the motor status based on the data transmitted in step 3, and issue control instructions based on the motor status.

[0011] Preferably, the motor operating parameters include the deviation rate between the stator flux observation value and the measured value, the dynamic harmonic distortion rate of the inverter switching frequency, the orthogonality error of the rotor pole position encoder sine and cosine signals, and the dynamic coupling degree of the vector control dq axis current, which are marked as Sf, Sd, Se and Sq respectively; the hydrological environment parameters include the water body conductivity gradient change rate, the colloidal particle size distribution density, the acoustic Doppler flow velocity profile and the water quality spectral absorption coefficient, which are marked as Cg, Cp, Ca and Cs respectively; the attitude dynamics parameters include the six-degree-of-freedom coupled motion angular acceleration, the zero-bias stability of the inertial navigation system, the attitude quaternion differential change rate and the real-time estimation value of the fluid dynamics damping coefficient, which are marked as Ms, Mb, Mq and Mh respectively; the operating efficiency parameters include the light intensity contrast of the dirt identification photoelectric sensor, the fluid dynamics efficiency loss coefficient, the robotic arm end effector posture error compensation amount and the energy consumption efficiency ratio, which are marked as Wc, Wh, We and Wr respectively.

[0012] Preferably, the stator flux observation value and the measured value deviation rate are collected using a Siemens SINAMICS S120 series high-precision Hall voltage sensor array (model 7MH3701), which is installed at both ends of the motor stator winding and continuously collected at a sampling rate of 10kHz during the entire operation of the motor. The sensor output signal is processed by an NI-9205 data acquisition card and then estimated and compared in real time by a flux observer based on the UKF algorithm. The inverter switching frequency dynamic harmonic distortion rate uses a Yokogawa WT5000 high-precision power analyzer, and the sampling port is connected to the output ends of each phase of the inverter. It is continuously monitored at a sampling rate of 100kHz during the entire operation of the underwater robot. The equipment transmits the real-time FFT analysis results to the control center via industrial Ethernet. The quadrature error of the rotor magnetic pole position encoder sine and cosine signals uses the Heidenhain ROD The 780C series high-precision photoelectric encoder is directly installed on the motor spindle end and cooperates with the REF6081 signal subdivider to achieve a resolution of 0.036 arc seconds. During the movement of the robot, the position data is transmitted in real time at a frequency of 1kHz through the RS-485 bus. The dynamic coupling degree of the dq axis current of the vector control uses the LEMHAH1DR 900-S high-precision closed-loop Hall current sensor, which is installed at the three-phase lead-out end of the motor and cooperates with the TITMS320F28379D DSP for real-time vector solution. The data is collected at a sampling rate of 20kHz during the whole process.

[0013] Preferably, the water conductivity gradient change rate adopts YSI EXO2 multi-parameter water quality analyzer, a probe is installed every 20 cm on the robot column to form a vertical profile array, and the data of each layer is collected at a frequency of 1 Hz during the cleaning and cruising process, and transmitted to the control center through the RS-485 bus. The particle size distribution density of the colloidal particles uses MalvernMastersizer 3000 laser particle size analyzer, and the peristaltic pump installed at the front end of the robot is used for flow-through online sampling, and a full spectrum scan is completed every 10 seconds. The acoustic Doppler velocity profile adopts RDI WorkHorse Sentinel 600kHz acoustic Doppler velocity profiler, which is fixedly installed on the bottom platform of the underwater robot and installed at a downward angle of 30°. The equipment adopts a four-beam transducer array with a working distance of 0.5-20 meters, and can collect water velocity profiles in 32 layers. During the robot cruising and fixed-point operation, data is continuously collected at a frequency of 2 Hz. The water quality spectral absorption coefficient uses TriOS The RAMSES-ACC-VIS high-precision underwater spectrophotometer is installed on the front detection arm of the robot through a waterproof bracket. It is equipped with a 3-meter-long quartz fiber probe. It can perform high-precision scanning in the 380-730nm band. During pollution cleaning operations and water quality monitoring, spectral scanning and collection are performed at a frequency of 0.5Hz.

[0014] Preferably, the six-degree-of-freedom coupled motion angular acceleration adopts Honeywell HG4930 high-precision MEMS inertial measurement unit, which is installed at the center of gravity of the robot. The sensor component includes a three-axis orthogonal accelerometer and a three-axis gyroscope with a sampling rate of 1kHz. A Northrop Grumman LN-200 high-performance fiber optic gyroscope is used as a reference calibration. The zero bias stability of the inertial navigation system adopts KVH The DSP-1760 three-axis fiber optic gyroscope has a zero bias stability better than 0.05° / h and is installed in a sealed constant temperature cabin. The system is preheated for at least 4 hours before operation and continuously collects gyro data during operation. Temperature compensation is performed in real time with the temperature sensor array. The GPS / DVL data is fused through the Kalman filter algorithm for zero bias estimation and correction. The data sampling rate is 200Hz and is transmitted through the RS-422 interface. The attitude quaternion differential change rate uses the OxTSxNAV550 dual-antenna RTK-GNSS / INS integrated navigation system. The main antenna is installed on the top of the robot and the secondary antenna is installed at the tail for heading reference. The system includes an IMU, which can provide an attitude accuracy of 0.1°. The attitude quaternion is output at a frequency of 100Hz during the movement of the robot. The onboard computer uses the fourth-order Runge-Kutta method to calculate the quaternion differential in real time. The real-time estimate of the fluid dynamics damping coefficient uses the ATI The Omega160 high-precision six-dimensional force / torque sensor is installed at the root and end of the robotic arm. It collects force data at a sampling rate of 500Hz during operation and performs real-time estimation based on the flow field pressure sensor array data.

[0015] Preferably, the light intensity contrast of the dirt recognition photoelectric sensor adopts a FLIR Blackfly S high-speed industrial camera with an LED structured light source, which is installed at the end of the robotic arm. During the cleaning process, images are collected at 60fps, and image contrast data is processed in real time through FPGA. The fluid dynamics efficiency loss coefficient adopts a Kistler 4007 series high-precision pressure sensor array. 12 measuring points are evenly arranged at key positions of the robot shell, combined with the ABB 266 series differential pressure transmitter installed in the inlet and outlet waterways. During the entire process of underwater robot movement and cleaning operations, pressure data at each point is continuously collected at a sampling rate of 200Hz. All sensors are connected to the control center through the PROFIBUS-DP bus, and the ANSYS Fluent embedded module is used for real-time CFD calculation. The efficiency loss coefficient is calculated in combination with the preset standard working condition database, and the posture error compensation amount of the end effector of the robotic arm is calculated using Keyence The LK-G5000 laser displacement sensor is combined with a binocular vision system and installed near the end of the robotic arm. During the operation, position data is collected at a frequency of 200Hz to calculate the error compensation amount. The energy efficiency ratio uses a Yokogawa WT3000E power analyzer to monitor the total input power, and the E+H Promag 53P electromagnetic flowmeter is used to measure the treated water volume. The efficiency indicators are collected and calculated at a frequency of 1Hz throughout the cleaning process.

[0016] Preferably, the time alignment realizes the timestamp synchronization of the data through the IEEE 1588 protocol, and determines the time-aligned data as the target data area, divides the target data area into various sub-data areas according to the underwater operation depth, and marks them as 1, 2...n in sequence, the data quality control uses the 3σ principle to perform initial screening of abnormal values, and adopts the Hampel filter to detect outliers, when the missing data time is a short time, the linear interpolation interpolation method is adopted, when the missing data time is a medium time, the Kalman filter prediction interpolation method is adopted, when the missing data time is a long time, it is marked as invalid data, the environmental factor compensation is compensated by supplementing the temperature sensor array to compensate for the temperature sensitive device, and adding the water depth pressure sensor to correct the pressure influence, and the signal processing is used to reduce noise and filter the original signal.

[0017] Preferably, the short time in the data quality control is specifically a missing data time of less than 100ms, the medium time is specifically a missing data time of 100ms-1s, and the long time is specifically a missing data time of more than 1s. The signal processing is performed by filtering with a Butterworth low-pass filter with a cutoff frequency of 1 / 4 of the sampling rate, and denoising is performed by wavelet transform, specifically db4 wavelet, with 5 layers of decomposition.

[0018] Preferably, the step 3 analyzes the motor operating parameters by establishing a motor motion state evaluation function, which is specifically expressed as: , MOS i represents the motor motion state evaluation value of the ith sub-data area, Sf i It represents the deviation rate between the stator flux observation value and the measured value in the ith sub-data area, Sd i Indicates the dynamic harmonic distortion rate of the inverter switching frequency in the ith sub-data area, Se i Represents the quadrature error of the rotor magnetic pole position encoder sine and cosine signals in the i-th sub-data area, Sq i It represents the dynamic coupling degree of the vector-controlled dq-axis current in the i-th sub-data region, and Π is a constant.

[0019] Preferably, the step 3 analyzes the hydrological environment parameters by establishing a hydrological environment adaptation function, which is specifically expressed as: , HEA i represents the hydrological environment adaptability of the ith sub-data area, Cg i represents the rate of change of water conductivity gradient in the ith sub-data area, Cp i represents the colloidal particle size distribution density of the ith sub-data area, Ca i represents the acoustic Doppler velocity profile of the ith sub-data region, Cs i Represents the water quality spectral absorption coefficient of the i-th sub-data area.

[0020] Preferably, the step 3 analyzes the posture dynamics parameters by establishing a posture dynamics compensation function, which is specifically expressed as: , DPC i Indicates the attitude dynamics compensation value of the ith sub-data area, Ms i Represents the angular acceleration of the six-degree-of-freedom coupled motion of the i-th sub-data area, Mb i represents the zero bias stability of the inertial navigation system in the ith sub-data area, Mq i Indicates the differential change rate of the attitude quaternion of the i-th sub-data area, Mh i Represents the real-time estimated value of the fluid dynamic damping coefficient of the i-th sub-data region.

[0021] Preferably, the step 3 analyzes the operation performance parameters by establishing an operation performance evaluation function, which is specifically expressed as: ,OPE i represents the operational performance evaluation value of the ith sub-data area, Wc i represents the light intensity contrast of the dirt recognition photoelectric sensor in the i-th sub-data area, Wh i represents the fluid dynamics efficiency loss coefficient of the ith sub-data region, Wei Indicates the compensation amount of the end effector posture error of the manipulator in the ith sub-data area, Wr i Represents the energy efficiency ratio of the i-th sub-data area.

[0022] Preferably, the specific method of the motor control is:

[0023] A1: Preprocessing: Normalize the function values ​​transmitted in step 3 so that each function value is on the same scale;

[0024] A2: Normal state model: Calculate the mean μ and standard deviation σ of each evaluation function in step 3 under normal conditions, and define the normal range as [μ-kσ,μ+kσ], where k is an adjustment coefficient for measuring the range;

[0025] A3: Real-time detection: During the operation of the target robot, determine whether each evaluation function exceeds the normal range. When two or more evaluation functions exceed the normal range, the motor state is judged to be abnormal. When two evaluation functions exceed the normal range, it is marked as a slight abnormality. When three evaluation functions exceed the normal range, it is marked as a moderate abnormality. When four evaluation functions exceed the normal range, it is marked as a serious abnormality.

[0026] A4: Control instructions: When the motor status is normal, continue to run and keep monitoring. When the motor status is slightly abnormal, reduce the motor operating power to 80%, increase the data sampling frequency from the standard 1Hz to 5Hz, and send a yellow warning signal to the control center. When the motor status is moderately abnormal, immediately reduce the motor power to 60%, start the backup control mode, send an orange alarm to the control center, and start emergency log recording. When the motor status is seriously abnormal, switch the motor to the lowest power mode of 20%. When there is no improvement for 30 seconds, perform a safe shutdown, send a red alarm to the control center, and lock the current system status data.

[0027] Preferably, a motor control system for underwater pollution cleaning operations of an intelligent robot includes a database, a control center, and a user information terminal, and also includes a sensor deployment module, a data preprocessing module, a model analysis module, and a motor control module.

[0028] Preferably, the database includes all data texts of the motor control system of the intelligent robot for underwater pollution cleaning operations, the control center is used to control the information text instructions output by each module, and the user information terminal is an information output device for receiving the motor control system of the intelligent robot for underwater pollution cleaning operations.

[0029] Preferably, the sensor deployment module is used to deploy sensors on the target robot, to collect motor operation parameters, hydrological environment parameters, posture dynamics parameters and operation efficiency parameters of the target robot during underwater pollution cleaning operations, and to transmit the collected parameters to the data preprocessing module;

[0030] The data preprocessing module is used to preprocess the data transmitted by the sensor deployment module, including standardization, time alignment, data quality control, environmental factor compensation and signal processing, and transmit the preprocessed data to the model analysis module;

[0031] The model analysis module is used to establish a mathematical model to analyze the data transmitted by the data preprocessing module, and transmit the analysis results to the motor control module;

[0032] The motor control module is used to determine the motor state according to the data transmitted by the model analysis module, and issue a control instruction according to the motor state.

[0033] Technical effects and advantages of the present invention:

[0034] The present invention comprehensively deploys sensors through sensor deployment, which can collect various key parameters of the intelligent robot in underwater pollution cleaning operations in real time and accurately, including motor operation parameters, hydrological environment parameters, posture dynamics parameters and operation efficiency parameters. The acquisition of these parameters provides a solid foundation for subsequent data preprocessing and model analysis, thereby ensuring the accuracy and effectiveness of motor control; the data preprocessing process includes standardization, time alignment, data quality control, environmental factor compensation and signal processing, which jointly ensure the accuracy and consistency of the data. Through time alignment, it can be ensured that the data collected by different sensors are synchronized in time, which is convenient for subsequent analysis. Data quality control and environmental factor compensation further improve the reliability of the data and reduce errors and interference. Signal processing further improves the quality of the data through noise reduction and filtering. , providing more accurate information for subsequent analysis; by establishing a mathematical model to analyze the pre-processed data through model analysis, the laws and characteristics in the data can be deeply explored, so as to achieve an accurate evaluation of the motor state. The establishment of the motor motion state evaluation function, the hydrological environment adaptation function, the attitude dynamics compensation function and the operation efficiency evaluation function respectively evaluates the motor state from different angles, ensuring the accuracy and comprehensiveness of the evaluation results. The results of these model analyses provide a scientific basis for subsequent motor control, making the control instructions more accurate and effective; through motor control, according to the results of model analysis, the motor control module can judge the state of the motor in real time and issue corresponding control instructions according to the state. The motor control strategy integrates the traditional threshold method and statistical concepts, and can adapt to data fluctuations more flexibly, improving the robustness and stability of the control. By reducing the motor operating power, increasing the data sampling frequency, sending early warning signals and other measures, timely measures can be taken when the motor state is abnormal to avoid the occurrence or expansion of faults, thereby ensuring the smooth progress of underwater pollution cleaning operations. At the same time, the motor control module can also record the operating status and abnormal information of the motor, providing valuable data support for subsequent maintenance and optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 It is a schematic diagram of the structure of the method of the present invention.

[0036] Figure 2 It is a schematic diagram of the system structure of the present invention. DETAILED DESCRIPTION

[0037] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0038] refer to Figure 1 A motor control method for underwater pollution cleaning operation of an intelligent robot is shown, and the specific steps include:

[0039] Step 1: Sensor deployment: It is used to deploy sensors on the target robot to collect the motor operation parameters, hydrological environment parameters, posture dynamic parameters and operation efficiency parameters of the target robot's underwater pollution cleaning operation, and transmit the collected parameters to step 2.

[0040] The motor operation parameters include the deviation rate between the stator flux observation value and the measured value, the dynamic harmonic distortion rate of the inverter switching frequency, the orthogonality error of the rotor magnetic pole position encoder sine and cosine signals, and the dynamic coupling degree of the vector control dq axis current, which are marked as Sf, Sd, Se and Sq respectively. The hydrological environment parameters include the water body conductivity gradient change rate, the colloidal particle size distribution density, the acoustic Doppler flow velocity profile and the water quality spectral absorption coefficient, which are marked as Cg, Cp, Ca and Cs respectively. The attitude dynamics parameters include the six-degree-of-freedom coupled motion angular acceleration, the zero-bias stability of the inertial navigation system, the attitude quaternion differential change rate and the real-time estimation value of the fluid dynamics damping coefficient, which are marked as Ms, Mb, Mq and Mh respectively. The operation efficiency parameters include the light intensity contrast of the dirt identification photoelectric sensor, the fluid dynamics efficiency loss coefficient, the robotic arm end effector posture error compensation amount and the energy consumption efficiency ratio, which are marked as Wc, Wh, We and Wr respectively.

[0041] The stator flux observation value and the measured value deviation rate are collected by Siemens SINAMICS S120 series high-precision Hall voltage sensor array (model 7MH3701), which is installed at both ends of the motor stator winding. The sensor output signal is continuously collected at a sampling rate of 10kHz during the whole operation of the motor. After being processed by the NI-9205 data acquisition card, the sensor output signal is estimated and compared in real time by the flux observer based on the UKF algorithm. The dynamic harmonic distortion rate of the inverter switching frequency uses the Yokogawa WT5000 high-precision power analyzer, and the sampling port is connected to the output end of each phase of the inverter. It is continuously monitored at a sampling rate of 100kHz during the whole operation of the underwater robot. The equipment transmits the real-time FFT analysis results to the control center through industrial Ethernet. The quadrature error of the rotor magnetic pole position encoder sine and cosine signals adopts Heidenhain ROD The 780C series high-precision photoelectric encoder is directly installed on the motor spindle end and cooperates with the REF6081 signal subdivider to achieve a resolution of 0.036 arc seconds. During the robot movement, the position data is transmitted in real time at a frequency of 1kHz through the RS-485 bus. The dynamic coupling degree of the dq axis current of the vector control uses the LEM HAH1DR 900-S high-precision closed-loop Hall current sensor, which is installed at the three-phase lead-out end of the motor and cooperates with the TI TMS320F28379D DSP for real-time vector solution. The data is collected at a sampling rate of 20kHz during the entire process.

[0042] The water conductivity gradient change rate is measured by a YSI EXO2 multi-parameter water quality analyzer. A probe is installed every 20 cm on the robot column to form a vertical profile array. During the cleaning and cruising process, data of each layer is collected at a frequency of 1 Hz and transmitted to the control center via the RS-485 bus. The particle size distribution density of the colloidal particles is measured by a Malvern Mastersizer3000 laser particle size analyzer. The peristaltic pump installed at the front end of the robot is used for flow-through online sampling, and a full spectrum scan is completed every 10 seconds. The acoustic Doppler velocity profile is measured by an RDI WorkHorse Sentinel 600kHz acoustic Doppler velocity profiler, which is fixedly installed on the bottom platform of the underwater robot at a downward angle of 30°. The equipment uses a four-beam transducer array with a working distance of 0.5-20 meters, and can collect water velocity profiles in 32 layers. During the robot's cruising and fixed-point operation, data is continuously collected at a frequency of 2 Hz. The water quality spectral absorption coefficient is measured by TriOS The RAMSES-ACC-VIS high-precision underwater spectrophotometer is installed on the front detection arm of the robot through a waterproof bracket. It is equipped with a 3-meter-long quartz fiber probe. It can perform high-precision scanning in the 380-730nm band. During pollution cleaning operations and water quality monitoring, spectral scanning and collection are performed at a frequency of 0.5Hz.

[0043] The six-degree-of-freedom coupled motion angular acceleration uses Honeywell HG4930 high-precision MEMS inertial measurement unit, which is installed at the center of gravity of the robot. The sensor assembly includes a three-axis orthogonal accelerometer and a three-axis gyroscope with a sampling rate of 1kHz. The Northrop Grumman LN-200 high-performance fiber optic gyroscope is used as a reference calibration. The zero bias stability of the inertial navigation system uses KVH The DSP-1760 three-axis fiber optic gyroscope has a zero bias stability better than 0.05° / h and is installed in a sealed constant temperature cabin. The system is preheated for at least 4 hours before operation and continuously collects gyro data during operation. Temperature compensation is performed in real time with the temperature sensor array. The GPS / DVL data is fused through the Kalman filter algorithm for zero bias estimation and correction. The data sampling rate is 200Hz and is transmitted through the RS-422 interface. The attitude quaternion differential change rate uses the OxTSxNAV550 dual-antenna RTK-GNSS / INS integrated navigation system. The main antenna is installed on the top of the robot and the secondary antenna is installed at the tail for heading reference. The system includes an IMU, which can provide an attitude accuracy of 0.1°. The attitude quaternion is output at a frequency of 100Hz during the movement of the robot. The onboard computer uses the fourth-order Runge-Kutta method to calculate the quaternion differential in real time. The real-time estimate of the fluid dynamics damping coefficient uses the ATI The Omega160 high-precision six-dimensional force / torque sensor is installed at the root and end of the robotic arm. It collects force data at a sampling rate of 500Hz during operation and performs real-time estimation based on the flow field pressure sensor array data.

[0044] The light intensity contrast of the dirt recognition photoelectric sensor adopts FLIR Blackfly S high-speed industrial camera with LED structure light source, which is installed at the end of the robot arm. During the cleaning process, images are collected at 60fps, and image contrast data is processed in real time through FPGA. The fluid dynamics efficiency loss coefficient adopts Kistler 4007 series high-precision pressure sensor array. 12 measuring points are evenly arranged at key positions of the robot shell, combined with ABB 266 series differential pressure transmitters installed in the inlet and outlet waterways. During the movement and cleaning of the underwater robot, the pressure data of each point is continuously collected at a sampling rate of 200Hz. All sensors are connected to the control center through PROFIBUS-DP bus, and ANSYS Fluent embedded module is used for real-time CFD calculation. The efficiency loss coefficient is calculated in combination with the preset standard working condition database. The position error compensation of the end effector of the robot arm uses Keyence LK-G5000 laser displacement sensor with binocular vision system, which is installed near the end of the robot arm. During the operation, position data is collected at a frequency of 200Hz to calculate the error compensation. The energy efficiency ratio adopts Yokogawa WT3000E power analyzer to monitor the total input power, and cooperates with E+H Promag The 53P electromagnetic flowmeter measures the amount of treated water and collects and calculates efficiency indicators at a frequency of 1 Hz throughout the cleaning process.

[0045] Step 2: Data preprocessing: used to preprocess the data transmitted in step 1, including standardization, time alignment, data quality control, environmental factor compensation and signal processing, and transmit the preprocessed data to step 3.

[0046] The time alignment realizes the synchronization of the timestamp of the data through the IEEE 1588 protocol, and determines the time-aligned data as the target data area. The target data area is divided into sub-data areas according to the underwater operation depth, and marked as 1, 2...n in sequence. The data quality control uses the 3σ principle to perform initial screening of abnormal values, and adopts the Hampel filter to detect outliers. When the missing data time is a short time, the linear interpolation method is adopted. When the missing data time is a medium time, the Kalman filter prediction interpolation method is adopted. When the missing data time is a long time, it is marked as invalid data. The environmental factor compensation compensates for the temperature sensitive device by supplementing the temperature sensor array, and adds the water depth pressure sensor to correct the pressure influence. The signal processing is used to reduce noise and filter the original signal.

[0047] It should be specifically explained in this embodiment that the specific depth of the underwater operation depth division is determined according to the actual operation situation, and this embodiment does not make a specific limitation, and the first sub-data area represents the initial state of the target robot performing underwater pollution cleaning operations and does not participate in the actual analysis.

[0048] The short time in the data quality control is specifically the missing data time of less than 100ms, the medium time is specifically the missing data time of 100ms-1s, and the long time is specifically the missing data time of more than 1s. The signal processing is performed by filtering with a Butterworth low-pass filter, with a cutoff frequency of 1 / 4 of the sampling rate, and denoising is performed by wavelet transform, specifically db4 wavelet, with 5 layers of decomposition.

[0049] Step 3: Model analysis: used to establish a mathematical model to analyze the data transmitted in step 2, and transmit the analysis results to step 4.

[0050] The step 3 analyzes the motor operating parameters by establishing a motor motion state evaluation function, which is specifically expressed as: , MOS i represents the motor motion state evaluation value of the ith sub-data area, Sf i It represents the deviation rate between the stator flux observation value and the measured value in the ith sub-data area, Sd i Indicates the dynamic harmonic distortion rate of the inverter switching frequency in the ith sub-data area, Se i Represents the quadrature error of the rotor magnetic pole position encoder sine and cosine signals in the i-th sub-data area, Sq i It represents the dynamic coupling degree of the vector-controlled dq-axis current in the i-th sub-data region, and Π is a constant.

[0051] The motor motion state evaluation function uses a square sum formula to comprehensively consider the impact of flux deviation and harmonic distortion on the motor. The harmonic distortion caused by the inverter switching frequency is processed by a sine function to reflect its nonlinear impact on the motor state. An offset is added to the rotor pole position encoder error to avoid the situation where the denominator is zero. The vector control dq axis current coupling uses a sigmoid function to smoothly respond to changes in coupling.

[0052] The step 3 analyzes the hydrological environment parameters by establishing a hydrological environment adaptation function, which is specifically expressed as follows: , HEA i represents the hydrological environment adaptability of the ith sub-data area, Cg i represents the rate of change of water conductivity gradient in the ith sub-data area, Cp i represents the colloidal particle size distribution density of the ith sub-data area, Ca i represents the acoustic Doppler velocity profile of the ith sub-data region, Cs i Represents the water quality spectral absorption coefficient of the i-th sub-data area.

[0053] The hydrological environment adaptation function uses linear adjustment and logarithmic operation to process the particle size for the conductivity gradient to more reasonably express the environmental changes. The acoustic Doppler velocity introduces a hyperbolic tangent function to limit the flow changes within a reasonable range. The square of the spectral absorption coefficient suppresses the complexity of water quality and avoids the amplification of extreme values.

[0054] The step 3 analyzes the posture dynamics parameters by establishing a posture dynamics compensation function, which is specifically expressed as: , DPC i Indicates the attitude dynamics compensation value of the ith sub-data area, Ms i Represents the angular acceleration of the six-degree-of-freedom coupled motion of the i-th sub-data area, Mb i represents the zero bias stability of the inertial navigation system in the ith sub-data area, Mq i Indicates the differential change rate of the attitude quaternion of the i-th sub-data area, Mh i Represents the real-time estimated value of the fluid dynamic damping coefficient of the i-th sub-data region.

[0055] The attitude dynamics compensation function considers exponential decay of acceleration to provide fast dynamic response adjustment. The inertial navigation system uses square root to extract offset influence for dynamic compensation. The attitude quaternion change introduces square to emphasize the necessity of high-precision attitude adjustment. The absolute value of the damping coefficient is processed to avoid the influence of negative value in calculation.

[0056] The step 3 analyzes the operation performance parameters by establishing an operation performance evaluation function, which is specifically expressed as: ,OPE i represents the operational performance evaluation value of the ith sub-data area, Wc i represents the light intensity contrast of the dirt recognition photoelectric sensor in the i-th sub-data area, Wh i represents the fluid dynamics efficiency loss coefficient of the ith sub-data region, We i Indicates the compensation amount of the end effector posture error of the manipulator in the ith sub-data area, Wr i Represents the energy efficiency ratio of the i-th sub-data area.

[0057] The operation efficiency evaluation function introduces an exponential function to adjust the rapid response of the dirt recognition contrast. The hydrodynamic efficiency loss is linearly adjusted and squared to control its extreme impact on the operation efficiency. The end effector posture error uses logarithmic and sine functions to make the compensation more accurate. The energy consumption efficiency uses a sine function to make the energy consumption level adjustment more detailed.

[0058] Step 4: Motor control: Determine the motor status based on the data transmitted in step 3, and issue control instructions based on the motor status.

[0059] The specific method of the motor control is:

[0060] A1: Preprocessing: Normalize the function values ​​transmitted in step 3 so that each function value is on the same scale;

[0061] A2: Normal state model: Calculate the mean μ and standard deviation σ of each evaluation function in step 3 under normal conditions, and define the normal range as [μ-kσ,μ+kσ], where k is an adjustment coefficient for measuring the range;

[0062] The k is usually 2 or 3, and can also be adjusted according to actual needs;

[0063] A3: Real-time detection: During the operation of the target robot, determine whether each evaluation function exceeds the normal range. When two or more evaluation functions exceed the normal range, the motor state is judged to be abnormal. When two evaluation functions exceed the normal range, it is marked as a slight abnormality. When three evaluation functions exceed the normal range, it is marked as a moderate abnormality. When four evaluation functions exceed the normal range, it is marked as a serious abnormality.

[0064] A4: Control instructions: When the motor status is normal, continue to run and keep monitoring. When the motor status is slightly abnormal, reduce the motor operating power to 80%, increase the data sampling frequency from the standard 1Hz to 5Hz, and send a yellow warning signal to the control center. When the motor status is moderately abnormal, immediately reduce the motor power to 60%, start the backup control mode, send an orange alarm to the control center, and start emergency log recording. When the motor status is seriously abnormal, switch the motor to the lowest power mode of 20%. When there is no improvement for 30 seconds, perform a safe shutdown, send a red alarm to the control center, and lock the current system status data.

[0065] The step 4 combines the traditional threshold method and the statistical concept method, which can adapt to data fluctuations more flexibly. It does not require complex model training and can be executed through simple statistical calculations. It is suitable for resource-constrained environments. By adjusting the coefficient k and the threshold ratio that exceeds the range, it can flexibly adapt to different application scenarios and sensitivity to anomalies.

[0066] refer to Figure 2 A motor control system for underwater pollution cleaning operations of intelligent robots includes a database, a control center and a user information terminal, as well as a sensor deployment module, a data preprocessing module, a model analysis module and a motor control module.

[0067] The database includes all data texts of the motor control system of the intelligent robot underwater pollution cleaning operation, the control center is used to control the information text instructions output by each module, and the user information terminal is an information output device for receiving the motor control system of the intelligent robot underwater pollution cleaning operation.

[0068] The sensor deployment module is used to deploy sensors on the target robot, collect motor operation parameters, hydrological environment parameters, posture dynamics parameters and operation efficiency parameters of the target robot's underwater pollution cleaning operation, and transmit the collected parameters to the data preprocessing module;

[0069] The data preprocessing module is used to preprocess the data transmitted by the sensor deployment module, including standardization, time alignment, data quality control, environmental factor compensation and signal processing, and transmit the preprocessed data to the model analysis module;

[0070] The model analysis module is used to establish a mathematical model to analyze the data transmitted by the data preprocessing module, and transmit the analysis results to the motor control module;

[0071] The motor control module is used to determine the motor state according to the data transmitted by the model analysis module, and issue a control instruction according to the motor state.

[0072] The present invention comprehensively deploys sensors through sensor deployment, which can collect various key parameters of the intelligent robot in underwater pollution cleaning operations in real time and accurately, including motor operation parameters, hydrological environment parameters, posture dynamics parameters and operation efficiency parameters. The acquisition of these parameters provides a solid foundation for subsequent data preprocessing and model analysis, thereby ensuring the accuracy and effectiveness of motor control; the data preprocessing process includes standardization, time alignment, data quality control, environmental factor compensation and signal processing, which jointly ensure the accuracy and consistency of the data. Through time alignment, it can be ensured that the data collected by different sensors are synchronized in time, which is convenient for subsequent analysis. Data quality control and environmental factor compensation further improve the reliability of the data and reduce errors and interference. Signal processing further improves the quality of the data through noise reduction and filtering. , providing more accurate information for subsequent analysis; by establishing a mathematical model to analyze the pre-processed data through model analysis, the laws and characteristics in the data can be deeply explored, so as to achieve an accurate evaluation of the motor state. The establishment of the motor motion state evaluation function, the hydrological environment adaptation function, the attitude dynamics compensation function and the operation efficiency evaluation function respectively evaluates the motor state from different angles, ensuring the accuracy and comprehensiveness of the evaluation results. The results of these model analyses provide a scientific basis for subsequent motor control, making the control instructions more accurate and effective; through motor control, according to the results of model analysis, the motor control module can judge the state of the motor in real time and issue corresponding control instructions according to the state. The motor control strategy integrates the traditional threshold method and statistical concepts, and can adapt to data fluctuations more flexibly, improving the robustness and stability of the control. By reducing the motor operating power, increasing the data sampling frequency, sending early warning signals and other measures, timely measures can be taken when the motor state is abnormal to avoid the occurrence or expansion of faults, thereby ensuring the smooth progress of underwater pollution cleaning operations. At the same time, the motor control module can also record the operating status and abnormal information of the motor, providing valuable data support for subsequent maintenance and optimization.

[0073] Secondly: In the drawings of the embodiments disclosed in the present invention, only the structures related to the embodiments disclosed in the present invention are involved, and other structures can refer to the general design. In the absence of conflict, the same embodiment and different embodiments of the present invention can be combined with each other;

[0074] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A motor control method for underwater pollution cleaning operations of an intelligent robot, characterized in that: include: Step 1: Sensor deployment: used to deploy sensors on the target robot to collect motor operation parameters, hydrological environment parameters, posture dynamics parameters and operation efficiency parameters of the target robot's underwater pollution cleaning operation, and transmit the collected parameters to step 2; The motor operation parameters include the deviation rate between the stator flux observation value and the measured value, the dynamic harmonic distortion rate of the inverter switching frequency, the orthogonality error of the rotor magnetic pole position encoder sine and cosine signals, and the dynamic coupling degree of the vector control dq axis current, which are marked as Sf, Sd, Se, and Sq, respectively; the hydrological environment parameters include the water body conductivity gradient change rate, the colloidal particle size distribution density, the acoustic Doppler flow velocity profile, and the water quality spectrum absorption coefficient, which are marked as Cg, Cp, Ca, and Cs, respectively; the attitude dynamics parameters include the six-degree-of-freedom coupled motion angular acceleration, the zero-bias stability of the inertial navigation system, the attitude quaternion differential change rate, and the real-time estimation value of the fluid dynamics damping coefficient, which are marked as Ms, Mb, Mq, and Mh, respectively; the operation efficiency parameters include the light intensity contrast of the dirt identification photoelectric sensor, the fluid dynamics efficiency loss coefficient, the robot end effector posture error compensation amount, and the energy consumption efficiency ratio, which are marked as Wc, Wh, We, and Wr, respectively; Step 2: Data preprocessing: It is used to preprocess the data transmitted in step 1, including standardization, time alignment, data quality control, environmental factor compensation and signal processing, and transmit the preprocessed data to step 3; Step 3: Model analysis: used to establish a mathematical model to analyze the data transmitted in step 2, and transmit the analysis results to step 4; The step 3 analyzes the motor operating parameters by establishing a motor motion state evaluation function, which is specifically expressed as: , MOS i represents the motor motion state evaluation value of the ith sub-data area, Sf i It represents the deviation rate between the stator flux observation value and the measured value in the ith sub-data area, Sd i Indicates the dynamic harmonic distortion rate of the inverter switching frequency in the ith sub-data area, Se i Represents the quadrature error of the rotor magnetic pole position encoder sine and cosine signals in the i-th sub-data area, Sq i represents the dynamic coupling degree of the vector controlled dq axis current in the i-th sub-data area, and Π is a constant; The step 3 analyzes the hydrological environment parameters by establishing a hydrological environment adaptation function, which is specifically expressed as follows: , HEA i represents the hydrological environment adaptability of the ith sub-data area, Cg i represents the rate of change of water conductivity gradient in the ith sub-data area, Cp i represents the colloidal particle size distribution density of the ith sub-data area, Ca i represents the acoustic Doppler velocity profile of the ith sub-data region, Cs i represents the water quality spectral absorption coefficient of the i-th sub-data area; The step 3 analyzes the posture dynamics parameters by establishing a posture dynamics compensation function, which is specifically expressed as: , DPC i Indicates the attitude dynamics compensation value of the ith sub-data area, Ms i Represents the angular acceleration of the six-degree-of-freedom coupled motion of the i-th sub-data area, Mb i represents the zero bias stability of the inertial navigation system in the ith sub-data area, Mq i Indicates the differential change rate of the attitude quaternion of the i-th sub-data area, Mh i represents the real-time estimated value of the fluid dynamics damping coefficient of the ith sub-data region; The step 3 analyzes the operation performance parameters by establishing an operation performance evaluation function, which is specifically expressed as: ,OPE i represents the operational performance evaluation value of the ith sub-data area, Wc i represents the light intensity contrast of the dirt recognition photoelectric sensor in the i-th sub-data area, Wh i represents the fluid dynamics efficiency loss coefficient of the ith sub-data region, We i Indicates the compensation amount of the end effector posture error of the manipulator in the ith sub-data area, Wr i represents the energy efficiency ratio of the i-th sub-data area; Step 4: Motor control: Determine the motor status based on the data transmitted in step 3, and issue control instructions based on the motor status.

2. A motor control method for underwater pollution cleaning operation of an intelligent robot according to claim 1, characterized in that: The time alignment implements the synchronization of the timestamp of the data through the IEEE 1588 protocol, and determines the time-aligned data as the target data area. The target data area is divided into sub-data areas according to the underwater operation depth, and marked as 1, 2...n in sequence. The data quality control uses the 3σ principle to perform the initial screening of abnormal values, and adopts the Hampel filter to detect outliers. When the missing data time is short, the linear interpolation method is adopted. When the missing data time is medium, the Kalman filter prediction interpolation method is adopted. When the missing data time is long, it is marked as invalid data. The environmental factor compensation compensates for the temperature sensitive device by supplementing the temperature sensor array, and adds the water depth pressure sensor to correct the pressure influence. The signal processing is used to reduce noise and filter the original signal; The short time in the data quality control is specifically the missing data time of less than 100ms, the medium time is specifically the missing data time of 100ms-1s, and the long time is specifically the missing data time of more than 1s. The signal processing is performed by filtering with a Butterworth low-pass filter, with a cutoff frequency of 1 / 4 of the sampling rate, and denoising is performed by wavelet transform, specifically db4 wavelet, with 5 layers of decomposition.

3. The motor control method for underwater pollution cleaning operation of an intelligent robot according to claim 1 is characterized in that: The specific method of the motor control is: A1： Preprocessing: Normalize the function values ​​transmitted in step 3 so that each function value is on the same scale; A2: Normal state model: Calculate the mean μ and standard deviation σ of each evaluation function in step 3 under normal conditions, and define the normal range as [μ-kσ,μ+kσ], where k is an adjustment coefficient for measuring the range; A3: Real-time detection: During the operation of the target robot, determine whether each evaluation function exceeds the normal range. When two or more evaluation functions exceed the normal range, the motor state is judged to be abnormal. When two evaluation functions exceed the normal range, it is marked as a slight abnormality. When three evaluation functions exceed the normal range, it is marked as a moderate abnormality. When four evaluation functions exceed the normal range, it is marked as a serious abnormality. A4: Control instructions: When the motor status is normal, continue to run and keep monitoring. When the motor status is slightly abnormal, reduce the motor operating power to 80%, increase the data sampling frequency from the standard 1Hz to 5Hz, and send a yellow warning signal to the control center. When the motor status is moderately abnormal, immediately reduce the motor power to 60%, start the backup control mode, send an orange alarm to the control center, and start emergency log recording. When the motor status is seriously abnormal, switch the motor to the lowest power mode of 20%. When there is no improvement for 30 seconds, perform a safe shutdown, send a red alarm to the control center, and lock the current system status data.

4. A motor control system for underwater pollution cleaning operations of an intelligent robot, used to use the motor control method for underwater pollution cleaning operations of an intelligent robot as described in any one of claims 1 to 3, comprising a database, a control center and a user information terminal, characterized in that: It also includes a sensor deployment module, a data preprocessing module, a model analysis module and a motor control module.

Images