Multifunctional test system

Abstract

The invention relates to the technical field of underwater robot testing, in particular to a multifunctional testing system. Comprising a hydrodynamic test device, a data acquisition device and a performance analysis device, the hydrodynamic test device is used for providing test environments of water flows with different speeds, different directions and different intensities for the underwater robot; the data acquisition equipment is used for acquiring operation data of the underwater robot in the water flow and current water flow data when the underwater robot operates in the water flow; and the performance analysis equipment is used for identifying the operation data and the current water flow data and determining a performance index score corresponding to the underwater robot. Therefore, a natural water flow environment can be accurately reproduced, and the influence of special terrains on water flow is simulated. And the test result can highly reflect the operation performance of the robot in a natural water area. And the accuracy of the determined performance index score corresponding to the underwater robot is ensured. The test system for simulating various real underwater environments under laboratory conditions is realized.

Description

Technical Field

[0001] The invention relates to the technical field of underwater robot testing, in particular to a multifunctional testing system. Background Art

[0002] Underwater robots must be able to operate stably in complex hydrodynamic environments when performing marine scientific research, underwater exploration, rescue, and maintenance missions. Existing underwater robot designs are often initially designed using simulation software, followed by field testing in natural waters to verify the feasibility and effectiveness of the design. However, this approach has several issues:

[0003] 1. Uncontrollable environment: Natural water conditions such as flow rate, water temperature and water quality are difficult to control precisely, which makes experimental results difficult to replicate and verify, affecting the accuracy and reliability of the test.

[0004] 2. High cost: Testing in natural water bodies requires a lot of logistical support, such as ships, divers and high transportation costs, which significantly increases R&D costs.

[0005] 3. Time-consuming: The testing cycle is long, and each test requires a lot of time to prepare and execute, which limits the speed and efficiency of design iteration.

[0006] 4. Data collection limitations: In natural water bodies, data collection is subject to many limitations, such as light in the underwater environment, water flow noise and other factors, which may lead to inaccurate or incomplete data.

[0007] With the rapid development of underwater robot technology and the continuous expansion of its application fields, the requirements for underwater robot performance are also constantly increasing. Therefore, it is urgent to develop a test system that can simulate multiple real underwater environments under laboratory conditions to support efficient and accurate testing and design optimization of underwater robots. Summary of the Invention

[0008] In view of this, the present invention provides a multifunctional test system to solve the urgent need to develop a test system that can simulate a variety of real underwater environments under laboratory conditions to support efficient and accurate testing and design optimization of underwater robots.

[0009] In a first aspect, the present invention provides a multifunctional test system suitable for testing underwater robots. The multifunctional test system includes: a hydrodynamic test device, a data acquisition device, and a performance analysis device, wherein:

[0010] Hydrodynamic testing equipment, used to provide underwater robots with a test environment for water flows of different speeds, directions, and intensities;

[0011] Data acquisition equipment, used to collect the operation data of the underwater robot in the water flow and the current water flow data when the underwater robot is running in the water flow;

[0012] Performance analysis equipment is used to identify operating data and current water flow data to determine the corresponding performance index score of the underwater robot.

[0013] The multifunctional testing system provided in the embodiments of the present application includes a hydrodynamic testing device that provides an underwater robot with a test environment featuring water flows of varying speeds, directions, and intensities, accurately reproducing natural water flow environments and simulating the effects of unique terrain on water flow. This allows underwater robots to be tested in an environment that closely resembles a real-world ocean scene, ensuring that test results closely reflect the robot's performance in actual marine operations, thereby improving the accuracy and reliability of the test. This ensures environmental controllability for underwater robot testing and reduces the cost and time associated with testing in natural water bodies. The data acquisition device collects operational data of the underwater robot in the water flow, as well as current water flow data during the robot's operation, ensuring the accuracy and completeness of the collected data and avoiding the inaccurate or incomplete data collected during testing in natural water bodies. The performance analysis device identifies the operational data and current water flow data to determine the corresponding performance indicator score for the underwater robot, ensuring the accuracy of the determined performance indicator score for the underwater robot. This eliminates the need for pre-assessment of the underwater robot and helps R&D personnel optimize the design in advance, improving the robot's stability and reliability. A test system that simulates a variety of real underwater environments under laboratory conditions has been implemented to support efficient and accurate testing and design optimization of underwater robots.

[0014] In an optional embodiment, the hydrodynamic test equipment includes a hydrodynamic test pool, a water circulation device, an environmental simulation device, a water flow generating device, and a water flow parameter regulating device; wherein:

[0015] The hydrodynamic test pool is used to hold water flow and provide a water flow environment for underwater robots;

[0016] The water circulation device is connected to the hydrodynamic test pool and is used to control the circulation of water in the hydrodynamic test pool to form a water flow;

[0017] Environmental simulation equipment, used to adjust the water temperature, salinity, and pressure in the hydrodynamic test tank to simulate the real ocean environment and test the adaptability and performance of underwater robots in different ocean conditions;

[0018] A water flow parameter adjustment device is connected to the water flow generating device and is used to adjust the water flow generating device according to test requirements so that the water flow generating device generates water flows and vortices of different speeds, directions, and intensities;

[0019] The water flow generating device is used to generate water flows of different speeds, directions, and intensities as well as vortices of different intensities and sizes under the control of the water flow parameter regulating device.

[0020] The multifunctional testing system provided in the embodiments of the present application contains a hydrodynamic test tank containing water, providing a water flow environment for the underwater robot. This provides the underwater robot with a reusable, easily observable, and easily monitored water flow environment. The precisely designed tank structure effectively reduces unnecessary disturbances in the water flow, ensuring water flow stability during testing, which is crucial for accurately measuring the performance parameters of the underwater robot under specific water flow conditions. A water circulation device, connected to the hydrodynamic test tank, controls the circulation of water in the test tank, converting it into a water flow. This ensures continuous water circulation within the test tank, simulating the flow conditions of natural water bodies. The environmental simulation device regulates the water temperature, salinity, and pressure in the hydrodynamic test tank to simulate a real ocean environment, enabling testing of the underwater robot's adaptability and performance under different ocean conditions. By precisely adjusting the water temperature, salinity, and pressure, it can simulate a variety of marine environments, from shallow to deep sea, and from tropical to polar seas. This allows the underwater robot to withstand various extreme conditions in a laboratory environment. The water flow parameter adjustment device, connected to the water flow generation device, is used to adjust the water flow generation device according to experimental requirements, enabling it to generate water flows of varying speeds, directions, and intensities, as well as vortices of varying strengths and sizes. Under the control of the water flow parameter adjustment device, the water flow generation device is designed to generate water flows of varying speeds, directions, and intensities, as well as vortices of varying strengths and sizes. Working in tandem with the water flow parameter adjustment device, the water flow generation device and the water flow generation device can generate a wide variety of water flow conditions. Researchers can flexibly adjust the water flow speed, direction, intensity, and vortex characteristics to suit different research objectives. When studying the maneuverability of underwater robots in complex sea conditions, high-speed water flows with strong vortices can be generated to observe whether the robots can respond quickly and accurately to control commands and maintain a stable motion. When testing the energy efficiency of underwater robots, uniform water flows of varying speeds can be set to measure the robot's power output and energy consumption at different flow rates, providing data support for optimizing energy management systems. This highly customizable water flow generation capability greatly expands the application scope of the test system, meeting a wide range of experimental needs, from basic research to engineering applications.

[0021] In an optional embodiment, the water flow generating device comprises: a plurality of water flow ejectors, a wave generator and a vortex generator which can be independently controlled by the water flow parameter regulating device, wherein:

[0022] The water jet is used to adjust the jet angle and flow rate under the control of the water flow parameter regulating device to generate water flows of different speeds, directions and intensities;

[0023] A wave maker is used to generate waves of various wavelengths and heights under the control of a water flow parameter regulating device;

[0024] The vortex generator is used to generate vortices of different intensities and sizes under the control of the water flow parameter regulating device.

[0025] The multifunctional test system provided in the embodiment of the present application, the water jet, is used to adjust the jet angle and flow rate under the control of the water flow parameter adjustment device to generate water flows of different speeds, directions and intensities, thereby flexibly constructing multiple water flow conditions and accurately simulating local water flow characteristics.

[0026] A wave generator, controlled by a water flow parameter control device, generates waves of various wavelengths and heights, accurately reproducing these different wave characteristics. When testing an underwater robot's wind and wave resistance, the periodic ups and downs of ocean waves can be simulated to observe the robot's tossing and turning under the action of waves, as well as how it maintains stability through its balance system and attitude control mechanism. A vortex generator, controlled by a water flow parameter control device, generates vortices of varying intensities and sizes. Vortex generators are capable of simulating these complex vortices, which exist in the ocean. When testing an underwater robot's performance in a vortex environment, vortices of varying intensities and sizes can be generated to observe the robot's motion after entering the vortex, such as whether it is sucked into the vortex or whether it can escape autonomously.

[0027] In an optional embodiment, the data acquisition device includes: a monitoring device, a flow rate measurement device, a vortex structure calculation device, and a pressure monitoring device, wherein:

[0028] A monitoring device is used to photograph the underwater robot and generate multiple underwater robot images; based on each underwater robot image, the corresponding speed, acceleration, motion posture and running trajectory of the underwater robot are determined;

[0029] Flow rate measuring equipment, used to test the flow rate of water when the underwater robot is running in the water flow;

[0030] Vortex structure calculation equipment, used to identify the vortex structure of underwater robots when operating in water flow;

[0031] Pressure monitoring equipment is used to monitor the pressure changes below the underwater robot when the underwater robot is running in the water flow.

[0032] The multifunctional test system provided in the embodiment of the present application includes a monitoring device for photographing an underwater robot and generating multiple underwater robot images; based on each underwater robot image, the corresponding speed, acceleration, motion posture and running trajectory of the underwater robot are determined. The monitoring device can accurately record the position information of the robot at different times, calculate the deviation of its running trajectory from the preset trajectory, and thus evaluate the accuracy of the navigation system. By analyzing the posture changes of the robot in adjacent images, the acceleration and motion posture parameters can be accurately calculated, which helps to gain a deeper understanding of the dynamic performance of the robot during startup, acceleration, steering, etc., and provide accurate data support for optimizing its control algorithm. The flow rate measurement device is used to test the water flow velocity of the underwater robot when it is running in the water flow, thereby providing accurate water flow velocity data to support complex flow field research. The vortex structure calculation device is used to identify the vortex structure of the underwater robot when it is running in the water flow. Identify the vortex structure of the underwater robot when it is running in the water flow, including information such as the strength, size and position of the vortex. When testing the robot's performance in vortex areas, data provided by the vortex structure calculation device helps researchers analyze the forces exerted by the vortex on the robot, such as how the suction and torque generated by the vortex affect the robot's trajectory and posture stability. This data can be used to improve the robot's anti-vortex design, such as optimizing the shell shape to reduce vortex forces, thereby improving the robot's safety and reliability in complex water flow environments containing vortices. Pressure monitoring equipment is used to monitor pressure changes below the underwater robot as it operates in the water flow, providing a visual reflection of the force applied to the robot in the water flow.

[0033] In an optional embodiment, the monitoring device is used to perform target recognition on the underwater robot image, and crop the underwater robot image according to the recognition result to obtain the target image;

[0034] Input each target image into a preset motion posture recognition model;

[0035] Perform convolution operation by sliding at least one convolution kernel on the target image to extract basic features of the target image and obtain a feature map; basic features include edge features and corner features;

[0036] Normalize each feature map to obtain a normalized feature map;

[0037] In the multiple residual blocks corresponding to each residual block group, the normalized feature map is processed by convolution, batch normalization and ReLU activation function, and then added to the original input normalized feature map to form a residual connection feature map; each residual block group consists of multiple residual blocks. At different stages, the number of residual blocks and the number of channels of the feature map will vary;

[0038] Perform a global average pooling operation on the residual connection feature map of each channel, averaging the residual connection feature map of each channel into one value, thereby converting the residual connection feature map into a one-dimensional vector;

[0039] Map the one-dimensional vector to the required output dimension, the output dimension is 3, corresponding to the pitch angle, roll angle and yaw angle respectively;

[0040] Through linear transformation and activation function, the 3D feature vector is converted into the final motion posture.

[0041] The multifunctional test system provided in the embodiment of the present application inputs each target image into a preset motion posture recognition model; performs a convolution operation by sliding at least one convolution kernel on the target image to extract the basic features of the target image and obtain a feature map. These basic features are representative key information in the image. The edge features can clearly outline the outline of the underwater robot, and the corner features can accurately locate the key parts of the robot, such as the joints of the robotic arm, the installation position of the sensor, etc. This enables the model to quickly and accurately identify the main structure of the robot from the image, providing a solid foundation for subsequent posture analysis. Then, each feature map is normalized to obtain a normalized feature map;

[0042] In the multiple residual blocks corresponding to each residual block group, the normalized feature maps are processed through convolution, batch normalization, and ReLU activation functions before being added to the original input normalized feature maps to form residual connection feature maps. Each residual block group consists of multiple residual blocks, and the number of residual blocks and the number of channels in the feature maps vary at different stages. The use of residual block groups and residual connections is a major advantage of this model. In traditional deep neural networks, the vanishing gradient problem often makes the model difficult to train as the number of network layers increases. Residual connections, however, add the original input normalized feature maps to the feature maps processed by convolution and other methods, allowing information to be transferred more directly within the network, effectively alleviating the vanishing gradient problem. This enables the model to build a deeper network structure, thereby learning more complex and advanced features, improving the model's expressiveness and performance. Next, global average pooling is performed on the residual connection feature map of each channel, averaging the residual connection feature maps of each channel to a single value, thereby converting the residual connection feature map into a one-dimensional vector. This operation not only effectively reduces the feature dimensionality and the computational complexity of the model, but also preserves the key feature information of each channel. Compared to traditional fully connected layers, global average pooling avoids the introduction of a large number of parameters, reduces the risk of overfitting, and improves the model's generalization ability. Furthermore, the one-dimensional vector obtained after global average pooling is more concise and efficient in subsequent processing. The model can more quickly perform linear transformations and activation functions on it, thereby quickly obtaining the final motion pose. In underwater robot testing scenarios with high real-time requirements, this provides timely feedback on the robot's posture, providing strong support for the smooth progress of the experiment. Finally, the one-dimensional vector is mapped to the required output dimension of 3, corresponding to the pitch, roll, and yaw angles. Through linear transformations and activation functions, the three-dimensional feature vector is converted into the final motion pose. This design directly models the motion pose of the underwater robot. The robot's posture information in three-dimensional space can be accurately described, providing a key indicator for evaluating the robot's motion performance and control effectiveness. It also meets practical application requirements: The resulting motion posture can be directly applied to underwater robot testing and control.

[0043] In an optional embodiment, the flow rate measuring device calculates the water flow rate based on the following formula:

[0044]

[0045] Where: V is the water flow velocity, ΔF d is the Doppler shift; F0 is the frequency of the transmitted ultrasonic wave; C is the speed of sound moving in water; θ is the angle between the transmitted and received waves, where V represents the flow velocity in the x, y, and z directions respectively.

[0046] In the multifunctional test system provided in the embodiment of the present application, the flow rate measuring device calculates the water flow rate based on the following formula: The accuracy of the calculated water flow rate is guaranteed.

[0047] In an optional embodiment, the vortex structure calculation device is used to calculate the vortex structure based on the following formula:

[0048]

[0049] Where u is the velocity vector, is the velocity gradient tensor;

[0050]

[0051] Where S is the symmetric part of the velocity gradient tensor decomposition, and Ω is the antisymmetric part of the velocity gradient tensor decomposition;

[0052] J=S 2 +Ω 2

[0053] Where J is the tensor S 2 +Ω 2 eigenvalue of ; perform eigenvalue decomposition on J and obtain its three eigenvalues ​​λ1≥λ2≥λ3;

[0054] Identify vortex structures in the flow field based on the λ2 criterion;

[0055]

[0056] Where Q is the vortex strength. The vortex core area is located by extracting the isosurface where Q = constant. The vortex size of the vortex structure in each direction is determined based on the shape and range of the isosurface.

[0057] The multifunctional testing system provided in the embodiment of the present application ensures the accuracy of the determined vortex structure, and ensures the accuracy of the determined vortex intensity and vortex scale.

[0058] In an optional embodiment, the operation data includes the speed, acceleration, motion posture and operation trajectory of the underwater robot, and the current water flow data includes the current water flow velocity, current vortex size, current vortex intensity and pressure changes below the underwater robot; wherein:

[0059] Performance analysis equipment for obtaining initial water flow velocity, initial vortex size, and initial vortex intensity in a hydrodynamic test device before testing the underwater robot;

[0060] Determining a change in water flow conditions based on a relationship between an initial water flow velocity, an initial vortex size, and an initial vortex intensity and a current water flow velocity, a current vortex size, and a current vortex intensity;

[0061] Based on the changes in water flow conditions, speed, acceleration, motion posture, running trajectory and current pressure, the corresponding indicator evaluation score results of the underwater robot are determined; wherein the indicator evaluation score results include at least one result of the propulsion performance evaluation score, the posture control performance evaluation score, and the trajectory tracking performance evaluation score.

[0062] The multifunctional testing system and performance analysis equipment provided in embodiments of the present application are used to obtain the initial water flow velocity, initial vortex size, and initial vortex strength from a hydrodynamic testing device before testing an underwater robot. Based on the relationship between the initial water flow velocity, initial vortex size, and initial vortex strength and the current water flow velocity, current vortex size, and current vortex strength, the system determines changes in water flow conditions, thereby enabling a comprehensive assessment of the robot's performance based on these changes. Based on changes in water flow conditions, velocity, acceleration, motion posture, trajectory, and current pressure, the system determines the corresponding indicator evaluation score for the underwater robot, ensuring the accuracy of the determined indicator evaluation score. This provides a quantitative basis for optimizing the design of the underwater robot. If the propulsion performance evaluation score is low, optimization can be performed on the thruster structure and power system parameters. If the attitude control performance evaluation score is unsatisfactory, improvements can be made to the attitude sensor accuracy and control algorithm. By continuously referencing the performance analysis results, the overall performance of the robot is gradually improved. Furthermore, the performance analysis equipment continuously provides feedback for the development of underwater robot technology.

[0063] In an optional embodiment, the performance analysis device is used to input the changes in water flow conditions, speed, acceleration and pressure changes into a preset propulsion performance evaluation model;

[0064] Extract features of water flow condition changes, speed, and acceleration, and output propulsion efficiency and speed stability;

[0065] Output propulsion performance evaluation score based on propulsion efficiency and speed stability;

[0066] and / or

[0067] Extract the features of motion posture to obtain the posture angle change range, posture angle change rate and posture pattern features;

[0068] Input the attitude angle change range, attitude angle change rate and attitude mode characteristics into the preset attitude energy evaluation model and output the attitude control performance evaluation score;

[0069] and / or

[0070] The DTW algorithm is used to calculate the similarity distance between the actual running trajectory and the ideal trajectory, and the trajectory tracking performance evaluation score is determined based on the similarity distance.

[0071] The multifunctional testing system provided in the embodiments of the present application inputs changes in water flow conditions, velocity, acceleration, and pressure into a preset propulsion performance evaluation model; performs feature extraction on these changes, velocity, and acceleration, and outputs propulsion efficiency and speed stability, comprehensively considering both external water flow interference and the robot's own motion state during operation. This ensures the accuracy of the output propulsion efficiency and speed stability, and outputs a propulsion performance evaluation score based on these values, ensuring the accuracy of the output propulsion performance evaluation score. Furthermore, the system also extracts features from the motion posture to obtain the attitude angle variation range, attitude angle variation rate, and attitude pattern characteristics; these are input into a preset attitude performance evaluation model to output an attitude control performance evaluation score. The attitude angle variation range reflects the magnitude of the robot's attitude adjustment under different operating conditions, the attitude angle variation rate reflects the timeliness of attitude adjustments, and the attitude pattern characteristics reveal the robot's typical attitude performance under specific tasks or water flow conditions. These features comprehensively cover key aspects of attitude control and provide rich information for attitude control performance evaluation. Inputting these features into a pre-set attitude performance evaluation model accurately analyzes their correlation with attitude control performance. For example, in strongly swirling water, if the attitude angle changes beyond the normal range and the rate of change is too rapid, combined with abnormal attitude pattern characteristics, the model can determine that the attitude control system's control accuracy and stability are insufficient when dealing with complex water flow disturbances. This model then outputs an accurate attitude control performance evaluation score, providing clear guidance for improving the attitude control system. Alternatively, the DTW algorithm can be used to calculate the similarity distance between the actual and ideal trajectory, and based on this distance, a trajectory tracking performance evaluation score is determined. The DTW algorithm, which calculates the similarity distance between the actual and ideal trajectory, considers the time series characteristics of the trajectory and can effectively address trajectory expansion and offset issues along the time axis. In actual underwater robot operations, due to factors such as water flow disturbances, the trajectory may be delayed or advanced in time. The DTW algorithm accurately measures the difference between the actual and ideal trajectory in these complex situations. Compared to simple distance calculation methods, it more accurately reflects the authenticity of trajectory tracking and provides a quantitative assessment of trajectory tracking performance.

[0072] In an optional embodiment, the performance analysis device is further used to: monitor the mechanical, electrical, and control system states of the underwater robot, and obtain mechanical system state data, electrical system state data, and control system state data;

[0073] When the propulsion performance evaluation score, and / or the attitude control performance evaluation score, and / or the trajectory tracking performance evaluation score are lower than the corresponding preset evaluation score threshold, the mechanical system status data, the electrical system status data, and the control system status data are analyzed to determine the reasons why the propulsion performance evaluation score, and / or the attitude control performance evaluation score, and / or the trajectory tracking performance evaluation score are lower than the corresponding preset evaluation score threshold.

[0074] In the multifunctional test system provided in the embodiment of the present application, the performance analysis device is further used to: monitor the mechanical, electrical and control system status of the underwater robot, obtain mechanical system status data, electrical system status data and control system status data; when the propulsion performance evaluation score, and / or the attitude control performance evaluation score, and / or the trajectory tracking performance evaluation score are lower than the corresponding preset evaluation score threshold, the mechanical system status data, the electrical system status data and the control system status data are analyzed to determine the reason why the propulsion performance evaluation score, and / or the attitude control performance evaluation score, and / or the trajectory tracking performance evaluation score are lower than the corresponding preset evaluation score threshold. When the performance evaluation score is lower than the preset threshold, comprehensive and in-depth fault diagnosis can be achieved through comprehensive analysis of the multi-system status data. For example, if the propulsion performance evaluation score is low, the device can simultaneously analyze the wear of the propeller in the mechanical system, the power output stability of the motor in the electrical system, and the execution of the propulsion control algorithm in the control system to accurately find the specific cause of the poor propulsion performance, rather than being limited to a single system investigation, thereby improving the accuracy and efficiency of fault diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0076] Figure 1 is a schematic structural diagram of a multifunctional test system according to an embodiment of the present invention;

[0077] Figure 2 is a schematic structural diagram of another multifunctional test system according to an embodiment of the present invention;

[0078] Figure 3 is a structural diagram of another multifunctional test system according to an embodiment of the present invention;

[0079] Figure 4 FIG. 4 is a schematic structural diagram of another multifunctional test system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0080] To make the purpose, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are 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 those skilled in the art without making creative efforts shall fall within the scope of protection of the present invention.

[0081] like Figure 1 As shown, the present invention provides a multifunctional test system suitable for testing underwater robots. The multifunctional test system includes: a hydrodynamic test device 1, a data acquisition device 2, and a performance analysis device 3, wherein: the hydrodynamic test device 1 and the data acquisition device 2 are both communicatively connected to the performance analysis device 3, wherein:

[0082] Hydrodynamic testing equipment 1, used to provide the underwater robot with a test environment of water flows of different speeds, directions and intensities;

[0083] Data acquisition device 2, used to collect the operation data of the underwater robot in the water flow and the current water flow data when the underwater robot is running in the water flow;

[0084] The performance analysis device 3 is used to identify the operation data and the current water flow data and determine the performance index score corresponding to the underwater robot.

[0085] Specifically, the hydrodynamic testing equipment 1 provides the underwater robot with a test environment featuring water flows of varying speeds, directions, and intensities. The underwater robot can then conduct underwater tests within the hydrodynamic testing equipment 1. Furthermore, the multifunctional testing system's data acquisition device 2 can collect operational data and current water flow data while the underwater robot is conducting tests within the water flow. The performance analysis device 3 can identify these operational and current water flow data to determine a performance score for the underwater robot.

[0086] The multifunctional testing system provided in the embodiments of the present application includes a hydrodynamic testing device 1 that provides an underwater robot with a test environment featuring water flows of varying speeds, directions, and intensities, accurately reproducing natural water flow environments and simulating the effects of specific terrain on water flow. This allows underwater robots to be tested in an environment that closely resembles a real-world ocean scene, ensuring that test results closely reflect the robot's performance in actual marine operations, thereby improving the accuracy and reliability of the test. This ensures environmental controllability for underwater robot testing, reduces the cost and time associated with testing in natural water bodies, and saves time. The data acquisition device 2 collects operational data of the underwater robot in the water flow, as well as current water flow data during the underwater robot's operation, ensuring the accuracy and completeness of the collected data and avoiding the inaccurate or incomplete data collected during testing in natural water bodies. The performance analysis device 3 identifies the operational data and current water flow data to determine the corresponding performance indicator score for the underwater robot, ensuring the accuracy of the determined performance indicator score for the underwater robot. This eliminates the need for pre-assessment of the underwater robot and helps R&D personnel optimize the design in advance, improving the robot's stability and reliability. A test system that simulates a variety of real underwater environments under laboratory conditions has been implemented to support efficient and accurate testing and design optimization of underwater robots.

[0087] In an optional embodiment, as Figure 2 As shown, the hydrodynamic test equipment 1 includes a hydrodynamic test pool 11, a water circulation device 12, an environmental simulation device 13, a water flow generating device 14 and a water flow parameter regulating device 15; wherein: the water flow parameter regulating device 15 is communicatively connected to the water flow generating device.

[0088] A hydrodynamic test pool 11 is used to hold water and provide a water flow environment for the underwater robot;

[0089] The water circulation device 12 is connected to the hydrodynamic test tank 11 and is used to control the circulation of water in the hydrodynamic test tank 11 to form a water flow;

[0090] Environmental simulation equipment 13, used to adjust the water temperature, salinity and pressure in the hydrodynamic test tank 11 to simulate the real ocean environment to test the adaptability and performance of the underwater robot under different ocean conditions;

[0091] The water flow parameter adjustment device 15 is connected to the water flow generating device 14 and is used to adjust the water flow generating device 14 according to the test requirements so that the water flow generating device 14 generates water flows and vortices of different speeds, directions, and intensities;

[0092] The water flow generating device 14 is used to generate water flows of different speeds, directions, and intensities, as well as vortices of different intensities and sizes, under the control of the water flow parameter regulating device 15 .

[0093] Specifically, as a physical container, the hydrodynamic test pool 11 must be structurally designed to meet requirements such as stable water flow, no leakage, and easy observation. It is generally made of high-strength, corrosion-resistant materials, such as special alloys or tempered glass. Optionally, the hydrodynamic test pool 11 can adopt a modular splicing design, which allows the size and shape of the test pool to be flexibly adjusted according to different test scenarios. For example, when studying the motion performance of an underwater robot in a narrow channel, multiple small modules can be spliced ​​into a narrow and long channel; when testing the impact of a large-area water flow field on the robot, they can be spliced ​​into a spacious square or circular test pool. In addition, a transparent smart display area is set on the pool wall to display various water flow parameters such as flow rate, flow direction, pressure distribution, etc. in real time, making it convenient for researchers to observe and record. The pool wall is usually smoothed to reduce water flow resistance and turbulence. By connecting to the water circulation equipment 12, the test pool can accommodate circulating water, creating a stable operating environment for the underwater robot.

[0094] The core component of the water circulation equipment 12 is a water pump, which drives the water pump impeller to rotate through a motor to extract water from one end of the hydrodynamic test pool 11, transport it to the other end through a pipeline, and then return it to the hydrodynamic test pool 11, thereby realizing the circulation of water. Optionally, the water circulation equipment 12 can also include an electronic device that can use sensors to monitor the water flow rate at different positions in the hydrodynamic test pool 11 in real time, and automatically adjust the speed and valve opening of the water pump through a feedback control algorithm to maintain the stability and uniformity of the water flow rate. In addition, adding a water purification module to the water circulation system can filter, disinfect and adjust the water quality of the circulating water in real time, ensuring that the water quality is stable during the test and is not affected by impurities, microorganisms and chemical substances, thereby improving the reliability of the test results.

[0095] In the environmental simulation device 13, water temperature regulation typically uses a heating wire or a refrigeration device to raise or lower the temperature of the water in the hydrodynamic test tank 11 through the principle of heat exchange. When the heating wire is energized, it generates heat, raising the water temperature; the refrigeration device compresses the refrigerant, causing it to evaporate in the evaporator and absorb heat, thereby lowering the water temperature. Salinity is regulated by adding precisely measured salt substances to the water and using a stirring device to uniformly dissolve them. Pressure regulation utilizes a sealed test tank and a pressure pump to inject air or other gas into the hydrodynamic test tank 11 to increase the pressure within the tank and simulate the high-pressure environment of the deep sea. Sensors monitor the water temperature, salinity, and pressure in real time, and feed the data back to the control system to achieve precise control of environmental parameters.

[0096] Optionally, the hydrodynamic test pool 11 can include multiple test areas. The environmental simulation device 13 can independently control each test area. For example, it can simultaneously simulate shallow-water, high-temperature, low-salinity areas and deep-sea, low-temperature, high-salinity areas, allowing the underwater robot to traverse different environments in a single test, thereby more comprehensively testing its adaptability.

[0097] The water flow parameter adjustment device 15 is a computer-controlled intelligent system. It uses sensors to obtain operating status information from the water flow generating device 14, such as the injection angle and flow rate of the water jet 141, the wave parameters of the wave generator 142, and the vortex intensity and size of the vortex generator 143. Researchers enter the required water flow parameters on the user interface. Based on a preset algorithm, the system automatically calculates and generates control signals, which are sent to the actuators of the water flow generating device 14, such as the motor and hydraulic device. This adjusts the operating parameters of the water flow generating device 14, achieving precise control of parameters such as water flow speed, direction, intensity, and vortex.

[0098] Optionally, the water flow parameter adjustment device 15 can also automatically optimize the water flow parameters based on the underwater robot's real-time operating status. For example, if the robot experiences instability or lacks power under certain water flow conditions, the device automatically adjusts the water flow parameters to simulate a water flow environment more suitable for the robot's operation, helping researchers quickly identify performance bottlenecks and optimization directions.

[0099] In an optional embodiment, as Figure 3 As shown, the water flow generating device 14 includes: a plurality of water flow ejectors 141, a wave generator 142 and a vortex generator 143 which can be independently controlled by the water flow parameter regulating device 15, wherein:

[0100] The water jet 141 is used to adjust the jet angle and flow rate under the control of the water flow parameter adjustment device 15 to generate water flows of different speeds, directions and intensities;

[0101] The wave generator 142 is used to generate waves of various wavelengths and wave heights under the control of the water flow parameter adjustment device 15;

[0102] The vortex generator 143 is used to generate vortices of different intensities and sizes under the control of the water flow parameter adjustment device 15.

[0103] Specifically, the water jet 141 is usually composed of a nozzle, a flow regulating valve and an angle adjustment mechanism. During operation, water is ejected through the nozzle under pressure to form a water flow. The flow regulating valve controls the flow of water, thereby adjusting the intensity of the water flow; the angle adjustment mechanism is driven by a motor or hydraulic device to change the spray angle of the nozzle to achieve the generation of water flows in different directions. Multiple water jets 141 can be used in combination, and by independently controlling the flow and angle of each jet, complex multi-directional water flows can be generated. Optionally, the water jet 141 can also be a retractable and flexible jet that can quickly change shape according to different test requirements to generate more complex and diverse water flows.

[0104] Common wave generators 142 are mechanical and pneumatic. Mechanical wave generators 142 use a motor to drive a cam, crank, or other mechanical structure, causing a wave plate to move up and down or back and forth, pushing the water to generate waves. By adjusting the motor speed and the amplitude of the wave plate's movement, the wavelength and height of the waves can be controlled. Pneumatic wave generators 142 use compressed air released underwater to form a bubble curtain, disturbing the water to generate waves. Wave parameters are adjusted by controlling the flow and pressure of the compressed air. Optionally, the wave generator 142 can be retractable and flexible, allowing it to quickly change shape according to different experimental requirements, generating more complex and diverse waves.

[0105] Vortex generator 143 utilizes the principles of fluid dynamics, creating a vortex through a unique structural design. For example, spiral blades or specially shaped obstacles are placed in the water flow path. As the water flows through, the blades or obstacles act on it, causing it to rotate, forming vortices of varying intensities and sizes. The vortex parameters can be controlled by adjusting the angle and spacing of the blades or the shape and position of the obstacles.

[0106] Optionally, the water flow generating device 14 can automatically adjust operating parameters based on the real-time water flow conditions within the hydrodynamic test tank 11 to maintain the desired flow conditions. For example, when a localized water flow disturbance occurs within the test tank, the water jet 141 automatically adjusts the jet angle and flow rate to compensate for the disturbance and maintain a stable water flow field.

[0107] In the multifunctional testing system provided by the present embodiment, a hydrodynamic test tank 11 holds water, providing a flow environment for the underwater robot. This provides a reusable, easily observable, and monitorable water environment for the underwater robot. The precisely designed tank structure effectively reduces unnecessary disturbances in the water flow, ensuring stable flow during testing, which is crucial for accurately measuring the performance parameters of the underwater robot under specific flow conditions. A water circulation device 12, connected to the hydrodynamic test tank 11, controls the circulation of water within the test tank, converting it into a flow. This ensures continuous water circulation within the test tank, simulating the flow conditions of natural water. An environmental simulation device 13 regulates the water temperature, salinity, and pressure within the hydrodynamic test tank 11, simulating a real ocean environment to test the underwater robot's adaptability and performance under various ocean conditions. By precisely adjusting the water temperature, salinity, and pressure, a variety of marine environments, from shallow to deep sea, and from tropical to polar seas, can be simulated. This allows the underwater robot to withstand a variety of extreme conditions in a laboratory setting. The water flow parameter adjustment device 15 is connected to the water flow generation device 14 and is used to adjust the water flow generation device 14 according to test requirements, so that the water flow generation device 14 generates water flows of different speeds, directions, and intensities, as well as vortices. The water flow ejector 141 is used to adjust the injection angle and flow rate under the control of the water flow parameter adjustment device 15, generating water flows of different speeds, directions, and intensities. This allows for the flexible construction of multi-dimensional water flow conditions and the precise simulation of local water flow characteristics. The wave generator 142 is used to generate waves of various wavelengths and heights under the control of the water flow parameter adjustment device 15, and can accurately reproduce these different wave characteristics. When testing the wind and wave resistance of an underwater robot, the cyclical rise and fall of ocean waves can be simulated to observe the robot's turbulence under the action of waves and how its balance system and posture control mechanism maintain stability. The vortex generator 143 is used to generate vortices of different intensities and sizes under the control of the water flow parameter adjustment device 15. The ocean contains vortices of varying intensities and sizes, and the vortex generator 143 can simulate these complex vortices. When testing the performance of underwater robots in vortex environments, vortices of varying intensities and sizes can be generated to observe the robot's motion after entering a vortex, such as whether it is sucked into the vortex or whether it can escape on its own. The water flow parameter adjustment device 15 and the water flow generation device 14 work together to produce a wide variety of water flow conditions. Researchers can flexibly adjust the water flow speed, direction, intensity, and vortex characteristics to suit their research objectives.When studying the maneuverability of underwater robots in complex sea conditions, high-speed water flows with strong vortices can be generated to observe whether the robots can quickly and accurately respond to control commands and maintain a stable motion. When testing the energy efficiency of underwater robots, uniform water flows of varying speeds can be set to measure the robot's power output and energy consumption at these different flow rates, providing data support for optimizing energy management systems. This highly customizable water flow generation capability greatly expands the application range of the test system, meeting a wide range of testing needs, from basic research to engineering applications.

[0108] In an optional embodiment, as Figure 4 As shown, the data acquisition device 2 includes: a monitoring device 21 , a flow rate measuring device 22 , a vortex structure calculation device 23 and a pressure monitoring device 24 , wherein: the vortex structure calculation device 23 is communicatively connected to the flow rate measuring device 22 .

[0109] The monitoring device 21 is used to photograph the underwater robot and generate multiple underwater robot images; based on each underwater robot image, the corresponding speed, acceleration, motion posture and running trajectory of the underwater robot are determined;

[0110] A flow rate measuring device 22 is used to measure the flow rate of the underwater robot when it is running in the water flow;

[0111] a vortex structure calculation device 23 for identifying the vortex structure of the underwater robot when it is operating in a water flow;

[0112] The pressure monitoring device 24 is used to monitor the pressure changes below the underwater robot when the underwater robot is running in the water flow.

[0113] Specifically, the monitoring equipment 21 typically consists of multiple high-definition cameras strategically positioned at various locations within the hydrodynamic test tank 11 to ensure comprehensive imaging of the underwater robot from multiple angles. The camera's optical system focuses light reflected from the underwater robot onto an image sensor, which converts the optical signal into an electrical signal. This signal is then processed by analog-to-digital conversion and an image processing chip to generate a digitized image of the underwater robot.

[0114] Optionally, monitoring equipment 21 may include, in addition to traditional visible light cameras, infrared cameras and ultrasonic imaging equipment. Infrared cameras can capture thermal images of underwater robots in low-light or turbid water environments, complementing the limitations of visible light images. Ultrasonic imaging equipment can penetrate objects of a certain thickness and provide information about the underwater robot's internal structure or surrounding hidden objects. Multimodal image fusion algorithms can integrate information from different image types, providing more comprehensive and accurate data for determining the robot's motion parameters.

[0115] Optionally, the monitoring device 21 calculates the speed of the underwater robot based on the displacement and time interval of the matching feature points in adjacent image frames. Assuming that the displacement of the feature point is Δx within the time interval Δt, the speed v = ΔtΔx. In order to obtain more accurate speed information, the speeds of multiple feature points can be averaged. Then, by calculating the speed changes at adjacent moments, the acceleration of the underwater robot is obtained. Assuming that the speeds at times t1 and t2 are v1 and v2 respectively, the acceleration a = (v2-v1) / (t2-t1). Similarly, the accelerations of multiple feature points can be averaged to improve the accuracy of the calculation.

[0116] The monitoring device 21 determines the center of mass of the underwater robot in each image frame. This can be done by calculating the geometric center of the robot's outline in the image frame or using a center of mass calculation method based on feature points. The center of mass positions at different times are connected to obtain the underwater robot's trajectory. To eliminate noise and jitter in the trajectory, the calculated trajectory is smoothed. Methods such as moving average filtering and spline interpolation can be used to make the trajectory smoother and more continuous.

[0117] The flow rate measuring device 22 can be a Doppler flowmeter, which operates based on the Doppler effect. The instrument transmits an ultrasonic signal into the water flow. When the ultrasonic wave encounters suspended particles in the water flow or an area of ​​uneven density in the water flow itself, it is scattered. Due to the movement of the water flow, the frequency of the scattered ultrasonic signal changes, and this frequency change is proportional to the water flow velocity. By measuring the frequency difference between the transmitted and received ultrasonic signals, combined with known parameters such as the speed of sound and the installation angle of the instrument, the Doppler formula is used to calculate the flow velocity of the water flow in different directions. Optionally, the flow rate measuring device 22 can automatically adjust the transmission frequency and measurement range of the flow rate measuring device 22 by monitoring the range of change in the water flow velocity in real time. For example, when the water flow velocity is low, the measurement resolution is increased to improve the accuracy of low-speed water flow measurement; when the water flow velocity is high, the measurement range is extended to avoid data distortion caused by the instrument exceeding the measurement range.

[0118] The vortex structure calculation device 23 can first obtain the three-dimensional velocity field data of the water flow around the underwater robot through the flow velocity measurement device 22. Then, based on the concept of the velocity gradient tensor, the velocity field is analyzed. The velocity gradient tensor describes the rate of change of the water flow velocity in all directions in space. The velocity gradient tensor is decomposed into a symmetric part and an antisymmetric part, and the antisymmetric part is related to the vortex. By calculating the eigenvalues ​​and eigenvectors of the tensor, specific criteria are used to identify the vortex structure in the flow field.

[0119] The pressure monitoring device 24 primarily consists of a pressure sensor, typically mounted on the bottom of the underwater robot or at other key locations. Common pressure sensors include strain gauge, piezoresistive, and capacitive types. For example, when pressure acts on the sensor's elastic diaphragm, the diaphragm deforms, causing the resistance of the strain gauge attached to the diaphragm to change. By measuring this change in strain gauge resistance and using a pre-calibrated pressure-resistance curve, the pressure acting on the sensor is calculated.

[0120] A pressure monitoring network composed of multiple pressure sensors collects real-time pressure data from different parts of the underwater robot. This data is amplified and filtered by signal conditioning circuits, converted into digital signals, and then transmitted to a data processing center via wired or wireless communication. There, the collected pressure data is stored, analyzed, and displayed, allowing researchers to understand the pressure changes experienced by the underwater robot as it navigates the water.

[0121] The multifunctional testing system provided in the embodiment of the present application includes a monitoring device 21 for photographing an underwater robot and generating multiple underwater robot images. Based on each underwater robot image, the corresponding speed, acceleration, motion posture, and trajectory of the underwater robot are determined. The monitoring device 21 can accurately record the robot's position information at different times and calculate the deviation between its trajectory and the preset trajectory, thereby evaluating the accuracy of the navigation system. By analyzing the robot's posture changes in adjacent images, the acceleration and motion posture parameters can be accurately calculated, which helps to gain a deeper understanding of the robot's dynamic performance during startup, acceleration, and steering, and provides accurate data support for optimizing its control algorithm. The flow rate measurement device 22 is used to test the flow velocity of the underwater robot while it is operating in a water flow, thereby providing accurate flow velocity data to support complex flow field research. The vortex structure calculation device 23 is used to identify the vortex structure of the underwater robot while it is operating in a water flow. The vortex structure of the underwater robot while it is operating in a water flow is identified, including information such as the strength, size, and position of the vortex. When testing the robot's performance in vortex areas, data provided by the vortex structure calculation device 23 helps researchers analyze the forces exerted by the vortex on the robot, such as how the suction and torque generated by the vortex affect the robot's trajectory and posture stability. This data can be used to improve the robot's anti-vortex design, such as optimizing the outer shell shape to reduce vortex forces, thereby improving the robot's safety and reliability in complex water flow environments containing vortices. The pressure monitoring device 24 is used to monitor the pressure changes below the underwater robot as it operates in the water flow, thereby visually reflecting the force state of the robot in the water flow.

[0122] In an optional embodiment, the monitoring device 21 is used to perform target recognition on the underwater robot image, and crop the underwater robot image according to the recognition result to obtain the target image;

[0123] Input each target image into a preset motion posture recognition model;

[0124] Perform convolution operation by sliding at least one convolution kernel on the target image to extract basic features of the target image and obtain a feature map; basic features include edge features and corner features;

[0125] Normalize each feature map to obtain a normalized feature map;

[0126] In the multiple residual blocks corresponding to each residual block group, the normalized feature map is processed by convolution, batch normalization and ReLU activation function, and then added to the original input normalized feature map to form a residual connection feature map; each residual block group consists of multiple residual blocks. At different stages, the number of residual blocks and the number of channels of the feature map will vary;

[0127] Perform a global average pooling operation on the residual connection feature map of each channel, averaging the residual connection feature map of each channel into one value, thereby converting the residual connection feature map into a one-dimensional vector;

[0128] Map the one-dimensional vector to the required output dimension, the output dimension is 3, corresponding to the pitch angle, roll angle and yaw angle respectively;

[0129] Through linear transformation and activation function, the 3D feature vector is converted into the final motion posture.

[0130] Specifically, the monitoring device 21 can analyze the underwater robot image using a target recognition algorithm (e.g., a deep learning-based target detection algorithm such as YOLO or Faster R-CNN) to identify the underwater robot's position and range within the image. Based on the target recognition results, the monitoring device 21 then crops the portion containing the underwater robot from the original image to obtain the target image. This reduces interference from irrelevant background information and improves the accuracy and efficiency of subsequent gesture recognition.

[0131] Then, the monitoring device 21 inputs each target image into a preset motion posture recognition model. Use convolution kernels of different sizes to slide on the target image at the same time, and extract features from the local area of ​​the target image by multiplying and summing the corresponding elements. Then, by fusing these multi-scale features, a more comprehensive and richer image feature representation is obtained. Different convolution kernels can extract different types of features. In this model, edge features and corner features are mainly extracted. Edge features can help identify the outline of the underwater robot, while corner features help determine the key parts of the robot, such as the positions of joints and sensors. Convolution kernels of different sizes can capture features of different scales. For example, in a convolution layer, 3x3, 5x5 and 7x7 convolution kernels are used at the same time, and their output feature maps are spliced ​​or fused to improve the model's ability to capture features of different scales.

[0132] Next, the feature maps are normalized to adjust their numerical ranges so that they have similar scales and distributions, resulting in normalized feature maps. This can accelerate model training and improve model stability and generalization. A common normalization method is batch normalization, which normalizes feature maps by calculating the mean and variance of each batch.

[0133] In the multiple residual blocks corresponding to each residual block group, the normalized feature map is processed through convolution, batch normalization, and the ReLU activation function, and then added to the normalized feature map of the original input to form a residual connection. This structure can solve the gradient vanishing problem in deep neural networks, allowing the model to build deeper network structures and learn more complex features. Among them, different residual block groups will have different numbers of residual blocks and channels in feature maps at different stages. In the shallow layers of the model, there are fewer residual blocks and channels, and they mainly extract some simple basic features. As the network deepens, the number of residual blocks increases, and the number of channels gradually increases, which can extract more advanced and abstract features.

[0134] Next, the residual connection feature maps for each channel are averaged to a single value, converting the residual connection feature map into a one-dimensional vector. For a feature map with multiple channels, each channel's feature map is a two-dimensional matrix. By averaging all elements in this matrix, a scalar value is obtained. Combining the scalar values ​​of all channels converts the two-dimensional feature map into a one-dimensional vector. This operation reduces the number of model parameters and computational complexity while preserving the key feature information of each channel and avoiding overfitting.

[0135] Finally, a fully connected layer maps the one-dimensional vector to the desired output dimension. Here, the output dimension is three, corresponding to the underwater robot's pitch, roll, and yaw angles. Each neuron in the fully connected layer is connected to every element of the input vector. The input vector is converted to an output vector by learning weight parameters. The three-dimensional feature vector mapped by the fully connected layer is then processed through a linear transformation and an activation function to obtain the final motion pose. Linear transformations adjust the numerical range and scale of the feature vector, while activation functions introduce nonlinear factors, enabling the model to learn more complex mapping relationships. Common activation functions include Reluctant Unified Unit (ReLU) and Sigmoid.

[0136] The multifunctional test system provided in the embodiment of the present application inputs each target image into a preset motion posture recognition model; performs a convolution operation by sliding at least one convolution kernel on the target image to extract the basic features of the target image and obtain a feature map. These basic features are representative key information in the image. The edge features can clearly outline the outline of the underwater robot, and the corner features can accurately locate the key parts of the robot, such as the joints of the robotic arm, the installation position of the sensor, etc. This enables the model to quickly and accurately identify the main structure of the robot from the image, providing a solid foundation for subsequent posture analysis. Then, each feature map is normalized to obtain a normalized feature map;

[0137] In the multiple residual blocks corresponding to each residual block group, the normalized feature maps are processed through convolution, batch normalization, and ReLU activation functions before being added to the original input normalized feature maps to form residual connection feature maps. Each residual block group consists of multiple residual blocks, and the number of residual blocks and the number of channels in the feature maps vary at different stages. The use of residual block groups and residual connections is a major advantage of this model. In traditional deep neural networks, the vanishing gradient problem often makes the model difficult to train as the number of network layers increases. Residual connections, however, add the original input normalized feature maps to the feature maps processed by convolution and other methods, allowing information to be transferred more directly within the network, effectively alleviating the vanishing gradient problem. This enables the model to build a deeper network structure, thereby learning more complex and advanced features, improving the model's expressiveness and performance. Next, global average pooling is performed on the residual connection feature map of each channel, averaging the residual connection feature maps of each channel to a single value, thereby converting the residual connection feature map into a one-dimensional vector. This operation not only effectively reduces the feature dimensionality and the computational complexity of the model, but also preserves the key feature information of each channel. Compared to traditional fully connected layers, global average pooling avoids the introduction of a large number of parameters, reduces the risk of overfitting, and improves the model's generalization ability. Furthermore, the one-dimensional vector obtained after global average pooling is more concise and efficient in subsequent processing. The model can more quickly perform linear transformations and activation functions on it, thereby quickly obtaining the final motion pose. In underwater robot testing scenarios with high real-time requirements, this provides timely feedback on the robot's posture, providing strong support for the smooth progress of the experiment. Finally, the one-dimensional vector is mapped to the required output dimension of 3, corresponding to the pitch, roll, and yaw angles. Through linear transformations and activation functions, the three-dimensional feature vector is converted into the final motion pose. This design directly models the motion pose of the underwater robot. The robot's posture information in three-dimensional space can be accurately described, providing a key indicator for evaluating the robot's motion performance and control effectiveness. It also meets practical application requirements: The resulting motion posture can be directly applied to underwater robot testing and control.

[0138] In an optional embodiment, the flow rate measuring device 22 calculates the water flow rate based on the following formula:

[0139]

[0140] Where: V is the water flow velocity, ΔF d is the Doppler shift; F0 is the frequency of the transmitted ultrasonic wave; C is the speed of sound moving in water; θ is the angle between the transmitted and received waves, where V represents the flow velocity in the x, y, and z directions respectively.

[0141] In the multifunctional test system provided in the embodiment of the present application, the flow rate measuring device 22 calculates the water flow rate based on the following formula:

[0142] The accuracy of the calculated water flow rate is guaranteed.

[0143] In an optional embodiment, the vortex structure calculation device 23 is used to calculate the vortex structure based on the following formula:

[0144]

[0145] Where u is the velocity vector, is the velocity gradient tensor;

[0146]

[0147] Where S is the symmetric part of the velocity gradient tensor decomposition, and Ω is the antisymmetric part of the velocity gradient tensor decomposition;

[0148] J=S 2 +Ω 2

[0149] Where J is the tensor S 2 +Ω 2 eigenvalue of ; perform eigenvalue decomposition on J and obtain its three eigenvalues ​​λ1≥λ2≥λ3;

[0150] If λ2<0, the point belongs to the vortex core region.

[0151]

[0152] Where Q is the vortex strength. The vortex core area is located by extracting the isosurface where Q = constant. The vortex size of the vortex structure in each direction is determined based on the shape and range of the isosurface.

[0153] The multifunctional testing system provided in the embodiment of the present application ensures the accuracy of the determined vortex structure, and ensures the accuracy of the determined vortex intensity and vortex scale.

[0154] In an optional embodiment, the operation data includes the speed, acceleration, motion posture and operation trajectory of the underwater robot, and the current water flow data includes the current water flow velocity, current vortex size, current vortex intensity and pressure changes below the underwater robot; wherein:

[0155] Performance analysis equipment 3, used to obtain the initial water flow velocity, initial vortex size and initial vortex strength in the hydrodynamic testing equipment 1 before testing the underwater robot;

[0156] Determining a change in water flow conditions based on a relationship between an initial water flow velocity, an initial vortex size, and an initial vortex intensity and a current water flow velocity, a current vortex size, and a current vortex intensity;

[0157] Based on the changes in water flow conditions, speed, acceleration, motion posture, running trajectory and current pressure, the corresponding indicator evaluation score results of the underwater robot are determined; wherein the indicator evaluation score results include at least one result of the propulsion performance evaluation score, the posture control performance evaluation score, and the trajectory tracking performance evaluation score.

[0158] Specifically, the performance analysis device 3 obtains the initial water flow velocity, initial vortex size, and initial vortex strength from the hydrodynamic testing device 1 before testing the underwater robot. The initial water flow velocity is then compared with the current water flow velocity, the initial vortex size is compared with the current vortex size, and the initial vortex strength is compared with the current vortex strength. Changes in the water flow conditions are determined based on the comparison results.

[0159] In an optional embodiment, the performance analysis device 3 is used to input the changes in water flow conditions, speed, acceleration and pressure changes into a preset propulsion performance evaluation model;

[0160] Extract features of water flow condition changes, speed, and acceleration, and output propulsion efficiency and speed stability;

[0161] Output propulsion performance evaluation score based on propulsion efficiency and speed stability;

[0162] Specifically, the performance analysis device 3 inputs changes in water flow conditions, velocity, acceleration, and pressure into a preset propulsion performance evaluation model. The preset propulsion performance evaluation model extracts features from these changes, velocity, and acceleration, generating flow condition change characteristics, velocity characteristics, and acceleration characteristics. These flow condition change characteristics encompass changes in parameters such as flow velocity, vortex size, and vortex intensity. By analyzing the time-varying curve of water velocity, features such as the rate of change and the range of velocity fluctuation are extracted. For example, the average change in water velocity per unit time is calculated to reflect the rate of acceleration or deceleration of the water flow; the difference between the maximum and minimum flow velocity values ​​within a certain time period is calculated to reflect the degree of flow velocity fluctuation. For vortex size and intensity, features such as the changing trend of vortex size, the peak value of vortex intensity, and the frequency of occurrence can be extracted. For example, the changes in vortex size at different times and the number of high-intensity vortices occurring per unit time can be observed. Velocity features include average velocity, velocity variation range, and the frequency of velocity changes. The average speed reflects the overall speed of the robot's movement over a period of time; the speed variation range reflects the fluctuation of the robot's speed, which can be obtained by calculating the difference between the maximum and minimum speed values; the speed change frequency counts the number of times the speed changes significantly per unit time and is used to measure the stability of the robot's speed. Acceleration characteristics include average acceleration, acceleration variation range, and acceleration direction change characteristics. Average acceleration can reflect the average rate of change of the robot's speed over a period of time; acceleration variation range shows the amplitude of acceleration fluctuation by calculating the difference between the maximum and minimum acceleration values; acceleration direction change characteristics can be obtained by analyzing the direction changes of the acceleration vector at different times, such as counting the number of times the acceleration direction changes and the angle of each change.

[0163] The pre-set propulsion performance evaluation model calculates propulsion efficiency based on extracted characteristics of water flow conditions, velocity, and acceleration, combined with physical mechanics and the law of conservation of energy. For example, the propulsion efficiency is calculated by analyzing the energy required to overcome water resistance and maintain speed at different water speeds and comparing it with the total energy consumed by the propulsion system. If the robot can maintain a stable speed with less energy consumption as the water speed increases, it indicates high propulsion efficiency.

[0164] The preset propulsion performance evaluation model determines speed stability based on the extracted speed and acceleration characteristics. If the speed variation range is small, the frequency of variation is low, and the acceleration fluctuations are stable, the robot's speed stability is good. For example, if the robot's speed remains within a narrow range over a period of time, with few speed changes and smooth changes in acceleration magnitude and direction, it can be judged to have high speed stability.

[0165] Then, the preset propulsion performance evaluation model automatically adjusts the weights of propulsion efficiency and speed stability based on the robot's current mission. For example, when the robot is performing long-distance cruising missions, speed stability is crucial for saving energy and maintaining course. In this case, the weight of speed stability can be increased to 0.6, while the weight of propulsion efficiency can be reduced to 0.4. However, when performing missions requiring rapid response and flexible steering, propulsion efficiency is more critical for rapidly changing speed and direction. Accordingly, its weight is increased to 0.7, while the weight of speed stability is reduced to 0.3. Finally, propulsion efficiency and speed stability are multiplied by their corresponding weights to output a propulsion performance evaluation score. For example, if propulsion efficiency is considered more critical to propulsion performance, it can be assigned a higher weight (such as 0.6), while speed stability can be weighted to 0.4. Let propulsion efficiency be \(E\) and speed stability be \(S\), resulting in a propulsion performance evaluation score of \(P = 0.6E + 0.4S\). In this way, the two key indicators are integrated into a single comprehensive score that intuitively reflects the propulsion performance of the underwater robot.

[0166] and / or

[0167] Extract the features of motion posture to obtain the posture angle change range, posture angle change rate and posture pattern features;

[0168] The attitude angle change range, attitude angle change rate and attitude mode characteristics are input into the preset attitude energy evaluation model to output the attitude control performance evaluation score.

[0169] Specifically, the performance analysis device 3 identifies the underwater robot's motion posture corresponding to each frame of the underwater robot image and determines the range of the robot's corresponding posture angle variation. For example, within 100 frames of continuously acquired underwater robot image data, the pitch angle values ​​corresponding to each frame are recorded. If the maximum value is 30° and the minimum value is -10°, then the pitch angle variation range is 40°. By calculating the variation ranges of these three posture angles, the magnitude of the robot's posture changes in different directions can be intuitively understood.

[0170] Then, the performance analysis device 3 identifies the attitude angle change range and determines the attitude angle change rate. Taking the pitch angle as an example, assuming that the pitch angle changes from θ1 to θ2 within the time interval Δt, then the change rate of the pitch angle is In practice, by performing differential calculations on the attitude angle data at a series of consecutive time points, we can obtain the rate of change of each attitude angle at different moments. These rate of change data can reflect the timeliness and agility of the robot's attitude adjustment.

[0171] Furthermore, performance analysis device 3 identifies the range and rate of attitude angle change to determine attitude pattern characteristics. These characteristics primarily focus on the typical patterns of attitude changes exhibited by the robot during task execution. For example, during a fixed-point hovering task, the robot may maintain a relatively stable posture, with a small range and low rate of attitude angle change. However, during a turning task, the yaw angle exhibits significant and regular changes. By analyzing a large amount of historical operational data and combining it with task type labels, we can identify the robot's attitude pattern characteristics for different task scenarios. These characteristics are crucial for understanding the robot's attitude control performance in different tasks.

[0172] Performance analysis device 3 inputs the attitude angle change range, attitude angle change rate, and attitude pattern characteristics into a preset attitude performance evaluation model. The preset attitude performance evaluation model processes and analyzes the input data based on its learned mapping relationship, ultimately outputting a quantitative attitude control performance evaluation score. This score can be a value between 0 and 100, with higher values ​​indicating better attitude control performance. It can also be a normalized value to facilitate comparison and comprehensive evaluation between different models.

[0173] In an optional implementation, a pre-set posture energy assessment model can automatically adjust its structure based on different task types and environmental complexity. For example, when performing routine tasks in simple water flow environments, the model can adopt a relatively simple structure to reduce computational effort and improve assessment efficiency. However, when performing high-precision tasks in complex water flow environments, the model automatically increases the number of network layers or adjusts the number of neurons to better learn the relationship between complex posture control characteristics and performance. This adaptive adjustment can be achieved through a meta-learning algorithm, which selects the optimal structure and parameter configuration for the posture energy assessment model based on task and environmental information.

[0174] and / or

[0175] The DTW algorithm is used to calculate the similarity distance between the actual running trajectory and the ideal trajectory, and the trajectory tracking performance evaluation score is determined based on the similarity distance.

[0176] Specifically, suppose there are two trajectory sequences, one is the actual running trajectory sequence A={a1,a2,...,a m}, the other is the ideal trajectory sequence B={b1,b2,...,b n}, where m and n are the lengths of the two sequences, a i and b j are points in the trajectory.

[0177] The DTW algorithm constructs an m×n distance matrix D, where D(i,j) represents the point a in the actual trajectory.i and point b in the ideal trajectory j The distance between points is usually measured using Euclidean distance, i.e. D(i,j)=dist(a i ,b j ).

[0178] After obtaining the similarity distance, we need to determine the trajectory tracking performance evaluation score based on this distance. Usually, a threshold range is set for evaluation.

[0179] For example, if the similarity distance is less than a small threshold θ1, it means that the actual trajectory is very close to the ideal trajectory, the trajectory tracking performance is very good, and a high evaluation score, such as above 90, can be given. If the similarity distance is between θ1 and a larger threshold θ2, it means that the trajectory tracking performance is average, and a medium evaluation score, such as between 60 and 90, can be given. If the similarity distance is greater than θ2, it means that the trajectory tracking performance is poor, and the evaluation score may be below 60.

[0180] The multifunctional testing system and performance analysis device 3 provided in the embodiment of the present application are used to obtain the initial water flow velocity, initial vortex size and initial vortex strength in the hydrodynamic testing device 1 before testing the underwater robot; based on the relationship between the initial water flow velocity, initial vortex size and initial vortex strength and the current water flow velocity, current vortex size and current vortex strength, the change in water flow conditions is determined, so that the robot performance can be comprehensively evaluated according to the change in water flow conditions.

[0181] The changes in water flow conditions, velocity, acceleration, and pressure are input into a pre-set propulsion performance evaluation model. Feature extraction is performed on these changes in water flow conditions, velocity, and acceleration to output propulsion efficiency and speed stability, fully accounting for external water flow interference and the robot's own motion state during operation. The accuracy of the output propulsion efficiency and speed stability is ensured, and a propulsion performance evaluation score is output based on these values, ensuring the accuracy of the output propulsion performance evaluation score. Alternatively, feature extraction is performed on the motion posture to obtain the attitude angle variation range, attitude angle variation rate, and attitude pattern characteristics. These are then input into a pre-set attitude performance evaluation model to output an attitude control performance evaluation score. The attitude angle variation range reflects the magnitude of the robot's attitude adjustment under different operating conditions, the attitude angle variation rate reflects the timeliness of attitude adjustments, and the attitude pattern characteristics reveal the robot's typical attitude performance under specific tasks or water flow conditions. These features comprehensively cover key aspects of attitude control and provide rich information for attitude control performance evaluation. Inputting these features into the pre-set attitude performance evaluation model accurately analyzes the correlation between these features and attitude control performance. For example, in strong vortexes, if the attitude angle changes beyond the normal range and the rate of change is too rapid, combined with abnormal attitude pattern characteristics, the model can determine that the attitude control system's control accuracy and stability are insufficient when dealing with complex water flow disturbances. This model then outputs an accurate attitude control performance evaluation score, providing clear direction for improving the attitude control system. Alternatively, the DTW algorithm can be used to calculate the similarity distance between the actual and ideal trajectory, and based on this similarity distance, a trajectory tracking performance evaluation score is determined. The DTW algorithm, which calculates the similarity distance between the actual and ideal trajectory, considers the time series characteristics of the trajectory and can effectively address trajectory expansion and offset issues along the time axis. In the actual operation of underwater robots, their trajectories may be delayed or advanced in time due to factors such as water flow disturbances. The DTW algorithm can accurately measure the difference between the actual and ideal trajectories in such complex situations. Compared to simple distance calculation methods, it more accurately reflects the authenticity of trajectory tracking and provides a quantitative assessment of trajectory tracking performance.

[0182] In an optional embodiment, the performance analysis device 3 is further used to: monitor the mechanical, electrical and control system states of the underwater robot, and obtain mechanical system state data, electrical system state data and control system state data;

[0183] When the propulsion performance evaluation score, and / or the attitude control performance evaluation score, and / or the trajectory tracking performance evaluation score are lower than the corresponding preset evaluation score threshold, the mechanical system status data, the electrical system status data, and the control system status data are analyzed to determine the reasons why the propulsion performance evaluation score, and / or the attitude control performance evaluation score, and / or the trajectory tracking performance evaluation score are lower than the corresponding preset evaluation score threshold.

[0184] Specifically, the performance analysis device 3 may include various detachable sensors. These sensors are detachably mounted on key mechanical components of the underwater robot. By installing various sensors on key mechanical components of the underwater robot (such as propellers, joints, transmission devices, etc.), the mechanical system status data is obtained. For example, a vibration sensor is installed at the rotating shaft of the propeller to monitor the vibration of the propeller in real time during operation. Abnormal vibration may mean that the propeller blades are damaged, the bearings are worn, and other problems; a position sensor is installed at the joint to monitor the rotation angle and position of the joint to determine whether the joint is stuck or loose.

[0185] Furthermore, the performance analysis device 3 can also collect data on various aspects of the underwater robot's electrical system based on its communication connection with the underwater robot, including battery charge, voltage, current, motor speed, temperature, and the operating status of various components in the circuit. For example, a power sensor can be used to monitor the remaining battery charge in real time, while voltage and current sensors can be used to detect the voltage and current values ​​in the electrical circuit to determine whether the electrical system has faults such as short circuits or overloads. A thermistor can be used to monitor the temperature of the motor; excessively high temperatures may indicate excessive motor load or poor heat dissipation.

[0186] The performance analysis device 3 can also obtain control instructions corresponding to the target robot and, based on the communication connection with the underwater robot, monitor the target robot's control system's command execution, the accuracy of sensor feedback data, and the operating status of the control algorithm. For example, it monitors whether the control instructions issued by the control system are correctly executed and compares the deviation between the command set value and the actual feedback value; checks whether the data transmitted by sensors (such as attitude sensors and position sensors) to the control system is stable and accurate; if the data fluctuates abnormally, it may affect the decision-making of the control algorithm; and analyzes the computing resource usage and running time of the control algorithm during operation to determine whether the algorithm is running efficiently and stably.

[0187] The propulsion performance evaluation score, attitude control performance evaluation score, and trajectory tracking performance evaluation score are then compared with the corresponding preset evaluation score thresholds. These thresholds are set based on the underwater robot's design requirements, actual application scenarios, and previous empirical data. They are used to measure whether the robot meets the expected standards for various performance indicators.

[0188] Optionally, when a performance evaluation score falls below a threshold, a comprehensive analysis of the mechanical, electrical, and control system status data is performed. For example, a low propulsion performance evaluation score could be due to damaged propeller blades in the mechanical system, resulting in reduced propulsion efficiency; insufficient motor power supply in the electrical system, affecting propeller power output; or a deviation in the propulsion control algorithm in the control system, preventing it from properly adjusting propeller operating parameters based on actual water flow conditions and robot status. Through detailed investigation and analysis of each system's status data, the root cause of the low performance score can be identified.

[0189] Optionally, the performance analysis device 3 can also input the mechanical system status data, electrical system status data, and control system status data into the underwater robot diagnostic model. The underwater robot diagnostic model can automatically determine possible performance problems and their causes based on the input mechanical, electrical, and control system status data. For example, when a set of electrical system status data is input showing an abnormal increase in motor current and an increase in propeller vibration in the mechanical system, the intelligent diagnostic model can quickly determine that the motor may be overloaded due to excessive propeller load, thereby affecting propulsion performance. The underwater robot diagnostic model can be trained based on a large amount of collected underwater robot mechanical, electrical, and control system status data under different working conditions, as well as the corresponding performance evaluation score data and the root causes corresponding to the performance evaluation score data.

[0190] In the multifunctional test system provided in the embodiment of the present application, the performance analysis device 3 is further used to: monitor the mechanical, electrical, and control system states of the underwater robot, obtain mechanical system state data, electrical system state data, and control system state data; when the propulsion performance evaluation score, and / or the attitude control performance evaluation score, and / or the trajectory tracking performance evaluation score are lower than the corresponding preset evaluation score threshold, analyze the mechanical system state data, electrical system state data, and control system state data to determine the reason why the propulsion performance evaluation score, and / or the attitude control performance evaluation score, and / or the trajectory tracking performance evaluation score are lower than the corresponding preset evaluation score threshold. When the performance evaluation score is lower than the preset threshold, comprehensive and in-depth fault diagnosis can be achieved through comprehensive analysis of the multi-system state data. For example, if the propulsion performance evaluation score is low, the device can simultaneously analyze the wear of the propeller in the mechanical system, the power output stability of the motor in the electrical system, and the execution of the propulsion control algorithm in the control system to accurately identify the specific cause of poor propulsion performance, rather than being limited to single system troubleshooting, thereby improving the accuracy and efficiency of fault diagnosis.

[0191] Although the embodiments of the present invention have been described with reference to the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention. Such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A multifunctional test system, characterized in that: Suitable for testing underwater robots, the multifunctional testing system includes: hydrodynamic testing equipment, data acquisition equipment, and performance analysis equipment, wherein: The hydrodynamic test equipment is used to provide the underwater robot with a test environment of water flows of different speeds, directions and intensities; The data acquisition device is used to collect the operation data of the underwater robot in the water flow and the current water flow data when the underwater robot is running in the water flow; The performance analysis device is used to identify the operating data and the current water flow data to determine the performance indicator score corresponding to the underwater robot.

2. The multifunctional test system according to claim 1, characterized in that: The hydrodynamic test equipment includes a hydrodynamic test pool, a water circulation device, an environmental simulation device, a water flow generating device, and a water flow parameter regulating device; wherein: The hydrodynamic test pool is used to contain water flow and provide a water flow environment for the underwater robot; The water circulation device is connected to the hydrodynamic test tank and is used to control the circulation of water in the hydrodynamic test tank to form a water flow; The environmental simulation equipment is used to adjust the water temperature, salinity and pressure in the hydrodynamic test tank to simulate a real ocean environment to test the adaptability and performance of the underwater robot under different ocean conditions; The water flow parameter adjustment device is connected to the water flow generating device and is used to adjust the water flow generating device according to test requirements so that the water flow generating device generates water flows and vortices of different speeds, directions and intensities; The water flow generating device is used to generate water flows of different speeds, directions and intensities and vortices of different intensities and sizes under the control of the water flow parameter regulating device.

3. The multifunctional test system according to claim 2, characterized in that: The water flow generating device comprises: a plurality of water flow ejectors, wave generators and vortex generators which can be independently controlled by the water flow parameter regulating device, wherein: The water jet is used to adjust the jet angle and flow rate under the control of the water flow parameter adjustment device to generate water flows of different speeds, directions and intensities; The wave generator is used to generate waves of various wavelengths and wave heights under the control of the water flow parameter regulating device; The vortex generator is used to generate vortices of different intensities and sizes under the control of the water flow parameter regulating device.

4. The multifunctional test system according to claim 1, characterized in that: The data acquisition equipment includes: monitoring equipment, flow rate measurement equipment, vortex structure calculation equipment and pressure monitoring equipment, wherein: The monitoring device is used to photograph the underwater robot to generate a plurality of underwater robot images; based on each of the underwater robot images, determine the speed, acceleration, motion posture and running trajectory of the underwater robot; The flow rate measuring device is used to test the water flow rate when the underwater robot is running in the water flow; a vortex structure calculation device for identifying the vortex structure of the underwater robot when it is operating in a water flow; The pressure monitoring device is used to monitor the pressure changes below the underwater robot when the underwater robot is running in the water flow.

5. The multifunctional test system according to claim 4, characterized in that: The monitoring device is used to perform target recognition on the underwater robot image, and crop the underwater robot image according to the recognition result to obtain a target image; Inputting each of the target images into a preset motion posture recognition model; Performing a convolution operation on the target image by sliding at least one convolution kernel to extract basic features of the target image to obtain a feature map; The basic features include edge features and corner features; Normalizing each of the feature maps to obtain a normalized feature map; In the multiple residual blocks corresponding to each residual block group, the normalized feature map is processed by convolution, batch normalization and ReLU activation function, and then added to the original input normalized feature map to form a residual connection feature map; each residual block group is composed of multiple residual blocks, and the number of residual blocks and the number of channels of the feature map will vary at different stages; Perform a global average pooling operation on the residual connection feature map of each channel, averaging the residual connection feature map of each channel into one value, thereby converting the residual connection feature map into a one-dimensional vector; Mapping the one-dimensional vector to the required output dimension, where the output dimension is 3, corresponding to the pitch angle, roll angle, and yaw angle respectively; Through linear transformation and activation function, the 3D feature vector is converted into the final motion posture.

6. The multifunctional test system according to claim 4, characterized in that: The flow rate measuring device calculates the water flow rate based on the following formula: Where: V is the water flow velocity, ΔF d is the Doppler shift; F0 is the frequency of the transmitted ultrasonic wave; C is the speed of sound moving in water; θ is the angle between the transmitted and received waves, where V represents the flow velocity in the x, y, and z directions respectively.

7. The multifunctional test system according to claim 4, characterized in that: The vortex structure calculation device is used to calculate the vortex structure based on the following formula: Where u is the velocity vector, is the velocity gradient tensor; Where S is the symmetric part of the velocity gradient tensor decomposition, and Ω is the antisymmetric part of the velocity gradient tensor decomposition; J=S 2 +Oh 2 Where J is the tensor S 2 +Ω 2 The characteristic value of Perform eigenvalue decomposition on J and obtain its three eigenvalues ​​λ1≥λ2≥λ3; Identify vortex structures in the flow field based on the λ2 criterion; Wherein, Q is the vortex strength, and the vortex core region is located by extracting the isosurface where Q=constant, and the vortex size of the vortex structure in each direction is determined based on the shape and range of the isosurface.

8. The multifunctional testing system according to claim 1, characterized in that: The operation data includes the speed, acceleration, motion posture and operation trajectory of the underwater robot; the current water flow data includes the current water flow velocity, current vortex size, current vortex intensity and pressure changes below the underwater robot; wherein: The performance analysis device is used to obtain the initial water flow velocity, initial vortex size and initial vortex strength in the hydrodynamic testing device before testing the underwater robot; determining a change in water flow conditions based on a relationship between the initial water flow velocity, the initial vortex size, and the initial vortex intensity and the current water flow velocity, the current vortex size, and the current vortex intensity; Based on the changes in the water flow conditions, the speed, the acceleration, the motion posture, the running trajectory and the current pressure, the indicator evaluation score result corresponding to the underwater robot is determined; wherein, the indicator evaluation score result includes at least one result of the propulsion performance evaluation score, the posture control performance evaluation score, and the trajectory tracking performance evaluation score.

9. The multifunctional testing system according to claim 8, characterized in that: The performance analysis device is used to Inputting the water flow condition change, the speed, the acceleration, and the pressure change into a preset propulsion performance evaluation model; Extract features of the water flow condition change, the speed, and the acceleration, and output propulsion efficiency and speed stability; outputting the propulsion performance evaluation score according to the propulsion efficiency and the speed stability; and / or Extracting features of the motion posture to obtain a posture angle change range, a posture angle change rate, and posture pattern features; Inputting the attitude angle change range, the attitude angle change rate, and the attitude mode feature into a preset attitude energy evaluation model, and outputting the attitude control performance evaluation score; and / or The similarity distance between the actual running trajectory and the ideal trajectory is calculated using the DTW algorithm, and the trajectory tracking performance evaluation score is determined based on the similarity distance.

10. The multifunctional test system according to claim 9, characterized in that: The performance analysis device is further used to: monitor the mechanical, electrical and control system states of the underwater robot, and obtain mechanical system state data, electrical system state data and control system state data; When the propulsion performance evaluation score, and / or the attitude control performance evaluation score, and / or the trajectory tracking performance evaluation score are lower than the corresponding preset evaluation score threshold, the mechanical system status data, the electrical system status data, and the control system status data are analyzed to determine the reason why the propulsion performance evaluation score, and / or the attitude control performance evaluation score, and / or the trajectory tracking performance evaluation score are lower than the corresponding preset evaluation score threshold.

Images