An obstacle avoidance control method for a transformer oil-immersed internal inspection aerial vehicle

Abstract

This application relates to the field of obstacle avoidance control technology and proposes an obstacle avoidance control method for a transformer oil-immersed internal inspection vehicle. The method includes: acquiring scanning results from a forward multibeam ultrasonic array and the jet pressure of each piezoelectric jet unit during the inspection process of the internal inspection vehicle; constructing a dynamic occupancy grid map and identifying free grids; solving for the uncoupled theoretical thrust of each piezoelectric jet unit, calculating the coupling coefficient and effective thrust, generating the optimal inspection path composed of free grids, and calculating the combined expected resultant force, expected torque vector, and thrust command of the inspection; determining whether a piezoelectric jet unit has failed based on its jet pressure; if so, removing the failed piezoelectric jet unit, reallocating the thrust, and regenerating the thrust command to achieve obstacle avoidance control of the internal inspection vehicle. This application can improve the stability and accuracy of obstacle avoidance control for oil-immersed internal inspection vehicles.

Description

Technical Field

[0001] This application relates to the field of obstacle avoidance control technology, specifically to an obstacle avoidance control method for a transformer oil-immersed internal inspection vehicle. Background Technology

[0002] During the long-term operation of oil-immersed transformers, defects such as insulation aging, partial discharge, and foreign matter deposition are prone to occur in the windings, core, leads, tap changers, and inner walls of the tank. These defects directly affect the safety and stability of the power system. Therefore, it is necessary to conduct regular internal inspections of the transformers to identify potential faults. Generally, an oil-immersed internal inspection vehicle based on the piezoelectric jet driving principle is used, equipped with camera and ultrasonic flaw detection modules. By relying on the alternating bending vibration of the piezoelectric vibrator to change the volume of the jet cavity, a continuous jet thrust is formed in the transformer oil, enabling close-range, full-view real-time inspection of the transformer's interior.

[0003] However, the interior of a transformer tank is filled with windings, partitions, leads, and other components, creating a narrow space with a complex flow field. Oil-immersed internal inspection vehicles rely on multiple piezoelectric jet units for coordinated drive, and the thrust of each unit is subject to severe flow field coupling interference, easily leading to thrust deviation, poor navigation stability, and weak maneuverability. Furthermore, existing obstacle avoidance control algorithms for microrobots are mostly adapted to conventional environments such as air or clean water, failing to consider the special operating conditions of transformer oil medium characteristics and piezoelectric jet coupling interference. This results in obstacle avoidance control strategies that cannot match the vehicle's driving characteristics and the oil working environment, making it difficult for the vehicle to accurately and stably avoid obstacles and reliably reach inspection points, severely restricting the safety and efficiency of transformer internal inspection operations. Summary of the Invention

[0004] This application provides an obstacle avoidance control method for a transformer oil-immersed internal inspection vehicle, to solve the problem that obstacle avoidance control of an oil-immersed internal inspection vehicle in a confined oil space is difficult to effectively match the vehicle's driving characteristics and the oil environment, resulting in an inability to stably reach the inspection point. The specific technical solution adopted is as follows:

[0005] One embodiment of this application provides an obstacle avoidance control method for a transformer oil-immersed internal inspection vehicle, the method comprising the following steps:

[0006] During the inspection of the internal inspection vehicle, the scanning results of the forward multibeam ultrasonic array and the jet pressure of each piezoelectric jet unit are collected. Based on the scanning results, a dynamic occupied grid map is constructed and free grids are identified.

[0007] Solve for the uncoupled theoretical thrust of each piezoelectric jet unit. Based on the interference reaction force of each piezoelectric jet unit under full thrust, calculate the coupling coefficient between piezoelectric jet units and the effective thrust of each piezoelectric jet unit. Generate the optimal inspection path composed of free grids. Based on the deviation vector of adjacent free grids in the optimal inspection path, the heading angle and angular velocity of each free grid of the internal inspection vehicle in the optimal inspection path, calculate the combined expected resultant force and expected moment vector of the inspection. Combined with the coupling coefficient, calculate the thrust command of each piezoelectric jet unit.

[0008] The piezoelectric jet unit is determined based on the jet pressure of the piezoelectric jet unit. If it is, the failed piezoelectric jet unit is removed, the thrust is redistributed, and the thrust command is regenerated to achieve obstacle avoidance control of the internal inspection vehicle.

[0009] Furthermore, the method for constructing the dynamically occupied grid map is as follows:

[0010] The scanning results of the forward multibeam ultrasound array include the azimuth angle, relative distance and echo intensity in each beam direction. Based on the echo intensity, some beams are filtered out to obtain the set of echo beams of the detected obstacles. The spatial location information of the detected obstacles is obtained and clustered to obtain the set of spatial location information of the obstacles.

[0011] Based on the current spatial location information of the internal inspection vehicle and the spatial location information of obstacles, a dynamic occupancy grid map of the internal space of the fuel tank is constructed.

[0012] Furthermore, the free grid is defined as a grid in the grid map that dynamically occupies a grid with an occupancy probability less than a preset obstacle threshold.

[0013] Furthermore, the method for calculating the coupling coefficient between the piezoelectric jet units is as follows:

[0014] The vehicle was placed in the transformer working oil, and each piezoelectric jet unit was turned on individually to full thrust. The interference reaction force of the remaining piezoelectric jet units was collected. The ratio of the interference reaction force of the remaining piezoelectric jet units to the full thrust of the turned-on piezoelectric jet units was recorded as the coupling coefficient between the remaining piezoelectric jet units and the turned-on piezoelectric jet units.

[0015] Furthermore, the effective thrust of the piezoelectric jet unit is calculated as follows:

[0016] Any piezoelectric jet element is designated as the target piezoelectric jet element, and any piezoelectric jet element that is not the target piezoelectric jet element is designated as the interfering piezoelectric jet element. The product of the coupling coefficient between the target piezoelectric jet element and the interfering piezoelectric jet element and the theoretical thrust of the interfering piezoelectric jet element is designated as the influence thrust of the target piezoelectric jet element and the interfering piezoelectric jet element. The sum of the influence thrusts of all interfering piezoelectric jet elements and the target piezoelectric jet element is designated as the comprehensive influence thrust. The difference between the theoretical thrust of the target piezoelectric jet element and the comprehensive influence thrust is designated as the effective thrust of the target piezoelectric jet element.

[0017] Furthermore, the desired resultant force is the product of the pre-calibrated position scaling gain and the deviation vector of adjacent waypoints.

[0018] Furthermore, the formula for calculating the desired torque vector is:

[0019]

[0020] in, This represents the desired torque vector along the Z-axis. This indicates the pre-calibrated heading proportional gain; This represents the pre-calibrated heading differential gain; This represents the desired heading angle, which is the direction of the expected resultant force of the vehicle relative to the vehicle's body coordinate system. The included angle of the axis; Indicates the current heading angle of the aircraft; This represents the angular velocity of the vehicle about the vertical z-axis.

[0021] Furthermore, the method for obtaining the thrust command of the piezoelectric jet unit is as follows:

[0022] An allocation matrix is ​​constructed based on the installation position vector and nozzle direction vector of the piezoelectric jet unit on the internal inspection vehicle. A coupling matrix is ​​constructed based on the coupling coefficient between the piezoelectric jet units. The product of the allocation matrix and the coupling matrix is ​​denoted as the effective allocation matrix.

[0023] The ratio of the desired torque vector to the characteristic length is denoted as the equivalent torque. Based on the combined desired resultant force and the equivalent torque, the command vector is constructed.

[0024] The thrust command vector for each piezoelectric jet element is calculated using the following formula:

[0025]

[0026] in, This represents the thrust command column vector for all piezoelectric jet units, where each element in the thrust command column vector represents the thrust command that each piezoelectric jet unit should execute. Represents the damped least-squares pseudo-inverse of the effective allocation matrix; This indicates the effective thrust of the piezoelectric jet unit; This represents the element-wise limiting function.

[0027] Furthermore, the specific method for determining whether a piezoelectric jet unit has failed based on the jet pressure of the piezoelectric jet unit includes:

[0028] When the jet pressure of the piezoelectric jet unit is lower than the preset pressure threshold, the piezoelectric jet unit corresponding to the jet pressure is determined to be faulty.

[0029] Furthermore, the method for reallocating thrust is as follows:

[0030] The ratio of the number of failed piezoelectric jet units to the total number of piezoelectric jet units is denoted as the failure ratio. The magnitude of the expected resultant force is reduced and the thrust is redistributed according to the difference between the number 1 and the failure ratio.

[0031] The beneficial effects of this application are:

[0032] This application first constructs a dynamically occupied grid map based on the detected obstacle information and identifies free grids. To accurately characterize the theoretical thrust of each piezoelectric jet unit under no-flow-field-interference conditions, the uncoupled theoretical thrust of each piezoelectric jet unit is solved. Since the high-speed jet of the piezoelectric jet unit alters the flow field distribution at the nozzles of other piezoelectric jet units, the actual output thrust of the piezoelectric jet unit is significantly lower than the uncoupled theoretical thrust. The flow field disturbance intensity between piezoelectric jet units is calibrated, and the coupling coefficient between piezoelectric jet units and the thrust of each piezoelectric jet unit are obtained. The effective thrust of the piezoelectric jet unit is obtained, and thrust compensation is performed based on the coupling effect characteristics to obtain the thrust command of each piezoelectric jet unit, avoiding the problem of mismatch between obstacle avoidance command and actual thrust. Finally, the failed piezoelectric jet units are eliminated, the thrust is redistributed, and the thrust command is regenerated to realize the obstacle avoidance control of the internal inspection vehicle. This solves the problem that the obstacle avoidance control of the oil-immersed internal inspection vehicle in the confined oil space is difficult to effectively match the vehicle's driving characteristics and oil environment, and cannot stably reach the inspection point. This improves the stability and accuracy of the obstacle avoidance control of the oil-immersed internal inspection vehicle. Attached Figure Description

[0033] Figure 1 This is a schematic flowchart of an obstacle avoidance control method for a transformer oil-immersed internal inspection vehicle, provided as an embodiment of this application. Detailed Implementation

[0034] Please see Figure 1 The diagram illustrates a flowchart of an obstacle avoidance control method for a transformer oil-immersed internal inspection vehicle according to an embodiment of this application. The method includes the following steps:

[0035] Step S001: During the inspection of the internal inspection vehicle, collect the scanning results of the forward multibeam ultrasonic array and the jet pressure of each piezoelectric jet unit, construct a dynamic occupied grid map based on the scanning results, and identify free grids.

[0036] The transformer oil tank lacks external positioning assistance signals such as GPS, and the internal oil is turbid with extremely poor visibility. Therefore, the transformer oil-immersed internal inspection vehicle cannot obtain external positioning and global environmental information, relying only on its onboard sensing equipment to collect local environmental data. Furthermore, the densely packed metal structure inside the transformer oil tank causes multipath reflections and false echoes during ultrasonic detection operations, leading to obstacle detection errors and severely interfering with the vehicle's obstacle avoidance path planning and determination. To overcome these sensing deficiencies and improve obstacle detection accuracy, invalid echo signals are eliminated in addition to detecting the obstacle's location and distance, avoiding detection deviations caused by multipath reflections and ensuring the accuracy of obstacle avoidance detection.

[0037] In this embodiment, an oil-immersed piezoelectric jet micro-vehicle is selected from the oil-immersed internal inspection vehicle for obstacle avoidance control. The transformer oil is a naphthenic mineral insulating oil, and the internal inspection vehicle is equipped with 4 piezoelectric jet units.

[0038] During the inspection of the internal inspection vehicle, a forward multi-beam ultrasonic array is used to perform a fan-shaped scan of the inside of the fuel tank to obtain the azimuth angle, relative distance and echo intensity in each beam direction. The jet pressure of each piezoelectric jet unit is collected by a jet pressure sensor. In this embodiment, the detection distance range is set to 0.05 to 1.5 m, the beam angle is ±30° and the sampling frequency is greater than or equal to 10 Hz.

[0039] Furthermore, beams with echo intensities less than the echo detection threshold in all beam directions are filtered out to obtain a set of echo beams that have detected obstacles. The echo distance, azimuth, and elevation angle of each beam in the set of echo beams that have detected obstacles are processed using a spherical coordinate to rectangular coordinate transformation method to obtain the spatial location information of the detected obstacles. The DBSCAN algorithm is then used to cluster the spatial location information of the detected obstacles to obtain a set of spatial location information of the obstacles.

[0040] The echo detection threshold is determined using the CFAR constant false alarm rate method; the spatial position information is the result of positioning based on the coordinate system of the internal detection vehicle body; the spatial position information set of obstacles consists of the spatial position information of each detected obstacle, with each obstacle corresponding to a spatial position information set; the CFAR constant false alarm rate method, the spherical coordinate to rectangular coordinate transformation method, and the DBSCAN algorithm are all well-known technologies and will not be described in detail here; in this embodiment, the neighborhood radius of the DBSCAN algorithm is set to 0.05m, and the minimum number of samples is set to 3.

[0041] At this point, a set of spatial location information for the obstacles has been obtained.

[0042] When an internal inspection vehicle uses single-beam ultrasound for internal inspection, it can only acquire localized, sparse obstacle information accompanied by random noise. Furthermore, dead reckoning by the internal inspection vehicle under oil conditions suffers from cumulative drift errors, which can easily lead to problems such as missed obstacles, misjudgments, and environmental map misalignment.

[0043] The scanning results of the forward multibeam ultrasonic array are uniformly mapped to a 3D raster map, and the incremental accumulation of the scanning results is achieved using a log-odds method. This method can effectively suppress random noise and fill in observation blind spots by utilizing temporal redundancy characteristics, and output continuous and stable representations of free space and obstacle-occupied space. It can simultaneously support global path planning and local reactive obstacle avoidance control during the vehicle's inspection process, enabling path planning and obstacle avoidance control to operate collaboratively under a unified spatial representation. At the same time, the map is updated in real time by combining the vehicle's motion increments, accurately adapting to the complex operating conditions of dense obstacles and narrow navigation space inside transformer tanks.

[0044] Specifically, based on the current spatial position information of the internal inspection vehicle and the spatial position information of obstacles, a dynamic occupancy grid map of the internal space of the fuel tank is constructed, and the occupancy probability of each grid in the dynamic occupancy grid map is obtained. In this embodiment, the grid side length is set to 0.02m, and the internal space of the fuel tank is discretized into a three-dimensional grid. The initial occupancy of each grid is 0, indicating an unoccupied state. Each time an echo signal is received, the spatial position of the obstacle is transformed from the vehicle body coordinate system to the grid coordinate system, and the occupancy probability of the corresponding spatial position point falling into the grid is calculated, thereby updating in real time based on the changes in the environment during real-time detection. Grids with an occupancy probability greater than or equal to a preset obstacle threshold are recorded as obstacle grids, and grids with an occupancy probability less than the preset obstacle threshold are recorded as free grids.

[0045] In this embodiment, the obstacle threshold is set to 0.6; the dynamic occupation grid map is a three-dimensional grid map.

[0046] Among them, the free grid is the free space for navigation; the construction of the three-dimensional grid map and the calculation of the occupancy probability are well-known technologies and will not be elaborated further.

[0047] At this point, the free grid is obtained.

[0048] Step S002: Solve for the uncoupled theoretical thrust of each piezoelectric jet unit. Based on the interference reaction force experienced by the remaining piezoelectric jet units when each piezoelectric jet unit is in full thrust state, calculate the coupling coefficient between piezoelectric jet units and the effective thrust of each piezoelectric jet unit. Generate the optimal inspection path composed of free grids. According to the deviation vector of adjacent free grids in the optimal inspection path, the heading angle and angular velocity of each free grid of the internal inspection vehicle in the optimal inspection path, calculate the combined expected resultant force and expected moment vector of the inspection. Combined with the coupling coefficient, calculate the thrust command of each piezoelectric jet unit.

[0049] The obstacle avoidance planning and thrust distribution adjustment actions of the internal inspection vehicle are all executed by each piezoelectric jet unit. The output thrust of the piezoelectric jet unit is affected by the excitation frequency, driving voltage, and transformer oil density. Since all piezoelectric jet units are in the same oil flow field, there is a significant flow field coupling effect in the output thrust of each unit. Therefore, it is necessary to establish a vehicle motion constraint model adapted to actual working conditions by combining the execution characteristics of the piezoelectric jet units and the flow field coupling law.

[0050] First, the uncoupled theoretical thrust of each piezoelectric jet unit is solved using a momentum thrust model. This momentum thrust model is constructed by combining the thrust efficiency of the piezoelectric jet device, transformer oil density, nozzle volumetric flow rate, and jet velocity parameters. It can accurately characterize the theoretical thrust of each piezoelectric jet unit under conditions without flow field interference. The specific calculation method is a well-known technique and will not be elaborated further.

[0051] In actual motion control, the high-speed jet from the piezoelectric jet unit alters the flow field distribution at the nozzles of other piezoelectric jet units through flow field entrainment and disturbance, resulting in the actual output thrust of the piezoelectric jet unit being significantly lower than the theoretical thrust without coupling. Therefore, it is necessary to pre-calibrate the flow field disturbance intensity between piezoelectric jet units and obtain the coupling coefficient between them.

[0052] The process of pre-calibrating the coupling coefficient of the piezoelectric jet unit is as follows: the internal inspection vehicle is placed in the transformer working oil, and each piezoelectric jet unit is turned on individually to full thrust. The interference reaction force of the remaining piezoelectric jet units is collected by the micro-force sensor. The ratio of the interference reaction force of the remaining piezoelectric jet units to the full thrust of the turned-on piezoelectric jet unit is recorded as the coupling coefficient between the remaining piezoelectric jet units and the turned-on piezoelectric jet unit.

[0053] Since the spacing between piezoelectric jet micro-vehicle units is larger than the nozzle characteristic size, and the transformer oil has high viscosity damping and low jet Reynolds number, the flow field interaction between piezoelectric jet units is mainly due to the entrainment effect. Within the normal operating thrust range, the coupling effect of each piezoelectric jet unit is approximately linear, and the coupling interference characteristics under multi-unit cooperative operation can be characterized by linear superposition.

[0054] Furthermore, any piezoelectric jet element is designated as the target piezoelectric jet element, and any piezoelectric jet element that is not the target piezoelectric jet element is designated as the interfering piezoelectric jet element. The product of the coupling coefficient between the target piezoelectric jet element and the interfering piezoelectric jet element and the theoretical thrust of the interfering piezoelectric jet element is designated as the influence thrust of the target piezoelectric jet element and the interfering piezoelectric jet element. The sum of the influence thrusts of all interfering piezoelectric jet elements and the target piezoelectric jet element is designated as the comprehensive influence thrust. The difference between the theoretical thrust of the target piezoelectric jet element and the comprehensive influence thrust is designated as the effective thrust of the target piezoelectric jet element.

[0055] Subsequent thrust allocation based on effective thrust can compensate for the characteristics of coupling effects, thus avoiding the mismatch between obstacle avoidance commands and actual thrust.

[0056] When all piezoelectric jet units output full thrust in the same direction, the internal inspection vehicle obtains the maximum effective thrust; when the piezoelectric jet units on both sides of the internal inspection vehicle output reverse thrust, the internal inspection vehicle achieves a stationary turn; and the maximum thrust of the piezoelectric jet unit is less than or equal to the effective thrust execution constraint.

[0057] The internal structure of a transformer tank is complex and densely packed with obstacles. If only local obstacle avoidance identification and control strategies are relied upon, the internal inspection vehicle is prone to getting trapped in a local optimum and failing to reach the target inspection point smoothly. Therefore, before implementing local obstacle avoidance control, it is necessary to plan the complete passage path of the vehicle from its current position to the target inspection point based on a dynamically occupied grid map.

[0058] Specifically, starting from the current grid position of the aircraft and ending at the preset inspection point of the fuel tank winding connector, the optimal inspection path is planned using the RRT* sampling planning algorithm within the free grid of the dynamically occupied grid map. This yields a waypoint sequence for the optimal inspection path, where each waypoint is a free grid. During the optimal inspection path generation process, if the thrust required for navigation between adjacent waypoints exceeds the effective thrust output by the piezoelectric jet unit, the path is deemed unsuitable for execution and discarded. In this embodiment, the iteration termination condition of the RRT* sampling planning algorithm is set to 2000 sampling points. The RRT* sampling planning algorithm is a well-known technology and will not be described further.

[0059] By using the effective thrust as the maximum thrust constraint of the piezoelectric jet unit and introducing a path planning feasible domain screening condition, the global planning path is fully matched with the actual driving capability of the vehicle, thus fundamentally avoiding the planning failure problem where the theoretical path is feasible but cannot be actually executed due to limitations in oil conditions and thrust limits.

[0060] After obtaining the optimal inspection path, the coordinated thrust distribution of the piezoelectric jet units needs to be carried out. When the vehicle travels along the optimal global path, the desired combined force required for the internal inspection vehicle needs to be calculated in real time based on the positional deviation of adjacent waypoints. The desired combined force is the product of the position proportional gain and the deviation vector of adjacent waypoints. The position proportional gain is calibrated by the attenuation curve method.

[0061] Given that the aircraft is driven collaboratively by multiple piezoelectric jet units with flow field coupling characteristics, it is necessary to decompose the desired resultant force into independent thrust commands for each piezoelectric jet unit. Therefore, a damped pseudo-inverse thrust allocation algorithm is adopted to solve for the optimal thrust combination of all piezoelectric jet units while accurately meeting the overall force output requirements. Simultaneously, by combining a flow field coupling compensation mechanism and a unit failure tolerance mechanism, precise matching between obstacle avoidance control commands and the actual performance of the aircraft is achieved, ensuring stable navigation within complex oil-confined spaces.

[0062] First, determine the desired torque vector for the vehicle to perform inspections along the optimal inspection path.

[0063]

[0064] in, This represents the desired torque vector along the Z-axis. Indicates the heading proportional gain; The heading differential gain is represented by the heading proportional gain, and both the heading differential gain and the heading differential gain are calibrated using the attenuation curve method. The desired heading angle represents the direction of the combined desired force of the piezoelectric jet units on the vehicle. Since the nozzle direction of the piezoelectric jet units is fixed, the direction of the combined desired force of the piezoelectric jet units can only remain stable relative to the environment when the vehicle's attitude is aligned with the desired heading. The desired heading angle is the direction of the combined desired force of the vehicle relative to the vehicle's coordinate system. The included angle of the axis; This indicates the vehicle's current heading angle, which is collected by the vehicle's inertial measurement unit. This represents the angular velocity of the vehicle about the vertical z-axis, collected by the vehicle's inertial measurement unit.

[0065] It is important to understand that since the aircraft is passively stabilized in oil by buoyancy and center of gravity trim, roll and pitch are achieved without active torque control, the desired torque only needs to be calculated in the Z-axis direction.

[0066] During the operation of the aircraft, the piezoelectric jet units are affected by flow field coupling. Therefore, thrust distribution is performed with flow field coupling compensation. Specifically:

[0067] An allocation matrix is ​​constructed based on the installation position vector and nozzle direction vector of the piezoelectric jet unit on the internal inspection vehicle. The allocation matrix The middle element represents the mapping relationship between the unit thrust of each piezoelectric jet unit and the overall resultant force and torque of the spacecraft, and the allocation matrix. The first three rows correspond to the dimensionless processing results of the force mapping of each piezoelectric jet element on different axes, and the allocation matrix is ​​used. The last three rows correspond to the dimensionless processing results of the torque mapping of each piezoelectric jet unit on different axes. In this embodiment, the allocation matrix is ​​constructed based on the pseudo-inverse allocation method. The construction of the allocation matrix based on the pseudo-inverse allocation method is a well-known technique and will not be described in detail here.

[0068] Furthermore, a coupling matrix is ​​constructed based on the coupling coefficients between the piezoelectric jet units. The coupling matrix is phalanx, The coupling matrix represents the number of piezoelectric jet units. All diagonal elements of the coupling matrix are 1s, and the off-diagonal elements are the negatives of the coupling coefficients between the piezoelectric jet units. Merge into the allocation matrix In this context, the product of the allocation matrix and the coupling matrix is ​​denoted as the effective allocation matrix. The elements in the effective allocation matrix represent the actual contribution of each piezoelectric jet unit's unit thrust command to the overall resultant force / torque of the vehicle after considering flow field coupling.

[0069] Furthermore, since torque and force have different dimensions, the desired torque vector is converted into an equivalent torque according to the characteristic length. The characteristic length refers to the length reference required to convert the desired torque vector into an equivalent torque. The characteristic length is calibrated by the nominal lever arm from the nozzle of the adjacent piezoelectric jet unit of the aircraft to the center of mass, that is, determined by the measured geometric layout dimensions of the aircraft nozzle. Specifically, the ratio of the desired torque vector to the characteristic length is denoted as the equivalent torque.

[0070] Based on equivalent moments Combined with the expected force Constructed as a 6-dimensional instruction vector [ , Specifically, [ , , , , , ], Instruction vector transpose, Includes the resultant force components on the X, Y, and Z coordinate axes. , , ; Equivalent torques on the X, Y, and Z coordinate axes , , .

[0071] Based on the effective allocation matrix and command vector, the thrust command column vector of each piezoelectric jet unit is calculated using the following formula:

[0072]

[0073] in, This represents the thrust command column vector for all piezoelectric jet units, where each element in the thrust command column vector represents the thrust command that each piezoelectric jet unit should execute. Represents the damped least-squares pseudo-inverse of the effective allocation matrix; This indicates the effective thrust of the piezoelectric jet unit; This represents an element-wise limiting function that ensures all elements in the thrust command column vector are greater than or equal to 0 and less than or equal to 0. This is to avoid the problem of the actual resultant force and the expected resultant force being mismatched due to the inability to execute the over-limit thrust command.

[0074] The damped least-squares pseudo-inverse of the effective allocation matrix is ​​used to numerically stably solve for the optimal thrust combination that satisfies the resultant force and torque requirements even when the effective allocation matrix is ​​rank deficient or close to singular. It should be noted that this embodiment uses the damped least-squares form for pseudo-inverse calculation, which still satisfies the requirement of numerical stability when the effective allocation matrix is ​​rank deficient or close to singular. Moreover, the thrust combination obtained satisfies the resultant force requirement optimally in the least-squares sense while having the smallest norm, thereby reducing the mutual coupling interference between multiple jet units in the same oil flow field.

[0075] Compared to the traditional method of allocating thrust to each piezoelectric jet unit according to theoretical thrust, the above process considers the coupling effect between different piezoelectric jet units in the actual oil flow field for compensation, thereby reducing the error between the actual resultant force and the expected resultant force during the obstacle avoidance control process of the aircraft. Furthermore, it limits the thrust to avoid the problem of the actual resultant force and the expected resultant force being seriously mismatched due to the inability to execute over-limit thrust commands.

[0076] At this point, the thrust command for each piezoelectric jet unit is obtained.

[0077] Step S003: Determine whether the piezoelectric jet unit is faulty based on the jet pressure of the piezoelectric jet unit. If so, remove the faulty piezoelectric jet unit, redistribute the thrust, and regenerate the thrust command to realize the obstacle avoidance control of the internal inspection vehicle.

[0078] Furthermore, during the obstacle avoidance control of the aircraft, if any piezoelectric jet unit fails, the thrust distribution needs to be adjusted adaptively in a timely manner to prevent the aircraft from losing control. Specifically: during the effective thrust distribution process, health monitoring and fault-tolerant reconstruction of the aircraft's piezoelectric jet units are performed. The jet pressure of each piezoelectric jet unit is collected in real time through jet pressure sensors. When the jet pressure of a piezoelectric jet unit is lower than a preset pressure threshold, the piezoelectric jet unit corresponding to that pressure is determined to have failed. The pressure threshold is determined through bench calibration. Based on the thrust command currently received by the piezoelectric jet unit, its theoretical nominal jet pressure under the corresponding operating condition is calculated, and 80% of this theoretical nominal jet pressure is used as the current real-time preset dynamic pressure threshold for that piezoelectric jet unit. The failed piezoelectric jet unit is removed from the distribution matrix. Neutralization and Coupling Matrix The remaining effective piezoelectric jet units are then removed from the list and recalculated into an effective allocation matrix. Based on this recalculated matrix, the thrust command for each remaining effective piezoelectric jet unit is recalculated.

[0079] It should be noted that during the internal inspection of the transformer oil tank, it is necessary to avoid jamming or loss of control caused by actuator failure during the inspection of the aircraft. Therefore, by removing the failed piezoelectric jet units and reconstructing the pseudo-inverse, it is possible to ensure that the obstacle avoidance capability can be maintained and the aircraft can return safely under the condition of reduced maneuverability. If the effective allocation matrix after removal cannot cover the direction of the desired resultant force, for example, if the nozzle directions of the remaining piezoelectric jet units are coplanar, it is determined that the maneuverability of the aircraft under the planned optimal inspection path is not feasible, and the optimal inspection path is replanned.

[0080] Specifically, the product of the effective allocation matrix and the thrust command of the remaining effective piezoelectric jet units is denoted as the actual achievable resultant force vector. The actual achievable resultant force vector includes the force components and torque components of the overall resultant force corresponding to all piezoelectric jet units on different axes. Based on the force components of the overall resultant force on different axes in the actual achievable resultant force vector, the amplitude of the actual achievable resultant force is calculated. If the amplitude of the actual achievable resultant force is greater than the amplitude of the synthesized expected resultant force, the local maneuverability is determined to be infeasible, that is, the synthesized expected resultant force exceeds the achievable resultant force of the piezoelectric jet unit after considering coupling and actual health status. At this time, it is necessary to replan the optimal inspection path.

[0081] It should be noted that if the piezoelectric jet unit fails during obstacle avoidance control on the optimal inspection path, thrust needs to be redistributed at the location of the failure to ensure the vehicle reaches the target inspection point smoothly.

[0082] Specifically, considering the degree of maneuverability retention and safety margin of the remaining piezoelectric jet units after failure, the ratio of the number of failed piezoelectric jet units to the total number of piezoelectric jet units is denoted as the failure ratio. The amplitude of the expected resultant force is reduced according to the difference between the number 1 and the failure ratio, and the thrust is redistributed. By reducing the thrust requirement, the combined expected resultant force is brought back to the achievable range. If the determination result of local maneuverability is still obtained after reducing the amplitude of the expected resultant force and redistributing the thrust, the optimal inspection path is replanned starting from the current position of the vehicle. After obtaining the optimal inspection path, obstacle avoidance control continues.

[0083] It should be noted that during the thrust redistribution process when the piezoelectric jet unit fails, the constraints of the effective thrust are recalculated using the remaining piezoelectric jet units, and then the optimal inspection path is re-determined using the updated constraints, thereby ensuring that the re-planned optimal inspection path can still be executed under the degraded maneuverability.

[0084] Finally, based on the thrust-voltage characteristic curve and linear interpolation method pre-calibrated by bench calibration, the excitation voltage amplitude required for each piezoelectric jet unit to execute the thrust command is obtained according to the thrust command of the remaining effective piezoelectric jet units. The corresponding piezoelectric oscillator is driven by the driver to generate the desired obstacle avoidance maneuver corresponding to the excitation voltage amplitude.

[0085] The upper limit of the excitation voltage amplitude of the piezoelectric jet unit is set to 210V, and the excitation frequency is kept at the resonant frequency, i.e., 110Hz.

[0086] This completes the obstacle avoidance control for the transformer oil-immersed internal inspection vehicle.

Claims

1. A method for obstacle avoidance control of a transformer oil-immersed internal inspection vehicle, characterized in that, The method includes the following steps: During the inspection of the internal inspection vehicle, the scanning results of the forward multibeam ultrasonic array and the jet pressure of each piezoelectric jet unit are collected. Based on the scanning results, a dynamic occupied grid map is constructed and free grids are identified. Solve for the uncoupled theoretical thrust of each piezoelectric jet unit. Based on the interference reaction force of each piezoelectric jet unit under full thrust, calculate the coupling coefficient between piezoelectric jet units and the effective thrust of each piezoelectric jet unit. Generate the optimal inspection path composed of free grids. Based on the deviation vector of adjacent free grids in the optimal inspection path, the heading angle and angular velocity of each free grid of the internal inspection vehicle in the optimal inspection path, calculate the combined expected resultant force and expected moment vector of the inspection. Combined with the coupling coefficient, calculate the thrust command of each piezoelectric jet unit. The piezoelectric jet unit is determined based on the jet pressure of the piezoelectric jet unit. If it is, the failed piezoelectric jet unit is removed, the thrust is redistributed, and the thrust command is regenerated to achieve obstacle avoidance control of the internal inspection vehicle.

2. The obstacle avoidance control method for a transformer oil-immersed internal inspection vehicle according to claim 1, characterized in that, The method for constructing the dynamically occupied grid map is as follows: The scanning results of the forward multibeam ultrasound array include the azimuth angle, relative distance and echo intensity in each beam direction. Based on the echo intensity, some beams are filtered out to obtain the set of echo beams of the detected obstacles. The spatial location information of the detected obstacles is obtained and clustered to obtain the set of spatial location information of the obstacles. Based on the current spatial location information of the internal inspection vehicle and the spatial location information of obstacles, a dynamic occupancy grid map of the internal space of the fuel tank is constructed.

3. The obstacle avoidance control method for a transformer oil-immersed internal inspection vehicle according to claim 1, characterized in that, The free grid is defined as a grid in the grid map that has an occupancy probability less than a preset obstacle threshold.

4. The obstacle avoidance control method for a transformer oil-immersed internal inspection vehicle according to claim 1, characterized in that, The method for calculating the coupling coefficient between the piezoelectric jet units is as follows: The vehicle was placed in the transformer working oil, and each piezoelectric jet unit was turned on individually to full thrust. The interference reaction force of the remaining piezoelectric jet units was collected. The ratio of the interference reaction force of the remaining piezoelectric jet units to the full thrust of the turned-on piezoelectric jet units was recorded as the coupling coefficient between the remaining piezoelectric jet units and the turned-on piezoelectric jet units.

5. The obstacle avoidance control method for a transformer oil-immersed internal inspection vehicle according to claim 1, characterized in that, The effective thrust of the piezoelectric jet unit is calculated as follows: Any piezoelectric jet element is designated as the target piezoelectric jet element, and any piezoelectric jet element that is not the target piezoelectric jet element is designated as the interfering piezoelectric jet element. The product of the coupling coefficient between the target piezoelectric jet element and the interfering piezoelectric jet element and the theoretical thrust of the interfering piezoelectric jet element is designated as the influence thrust of the target piezoelectric jet element and the interfering piezoelectric jet element. The sum of the influence thrusts of all interfering piezoelectric jet elements and the target piezoelectric jet element is designated as the comprehensive influence thrust. The difference between the theoretical thrust of the target piezoelectric jet element and the comprehensive influence thrust is designated as the effective thrust of the target piezoelectric jet element.

6. The obstacle avoidance control method for a transformer oil-immersed internal inspection vehicle according to claim 1, characterized in that, The desired resultant force is the product of the pre-calibrated position scaling gain and the deviation vector of adjacent waypoints.

7. The obstacle avoidance control method for a transformer oil-immersed internal inspection vehicle according to claim 1, characterized in that, The formula for calculating the desired torque vector is: in, This represents the desired torque vector along the Z-axis. This indicates the pre-calibrated heading proportional gain; This represents the pre-calibrated heading differential gain; This represents the desired heading angle, which is the direction of the expected resultant force of the vehicle relative to the vehicle's body coordinate system. The included angle of the axis; Indicates the current heading angle of the aircraft; This represents the angular velocity of the vehicle about the vertical z-axis.

8. The obstacle avoidance control method for a transformer oil-immersed internal inspection vehicle according to claim 1, characterized in that, The method for obtaining the thrust command of the piezoelectric jet unit is as follows: An allocation matrix is ​​constructed based on the installation position vector and nozzle direction vector of the piezoelectric jet unit on the internal inspection vehicle. A coupling matrix is ​​constructed based on the coupling coefficient between the piezoelectric jet units. The product of the allocation matrix and the coupling matrix is ​​denoted as the effective allocation matrix. The ratio of the desired torque vector to the characteristic length is denoted as the equivalent torque. Based on the combined desired resultant force and the equivalent torque, the command vector is constructed. The thrust command vector for each piezoelectric jet element is calculated using the following formula: in, This represents the thrust command column vector for all piezoelectric jet units, where each element in the thrust command column vector represents the thrust command that each piezoelectric jet unit should execute. Represents the damped least-squares pseudo-inverse of the effective allocation matrix; This indicates the effective thrust of the piezoelectric jet unit; This represents the element-wise limiting function.

9. The obstacle avoidance control method for a transformer oil-immersed internal inspection vehicle according to claim 1, characterized in that, The specific method for determining whether a piezoelectric jet unit has failed based on the jet pressure of the piezoelectric jet unit includes: When the jet pressure of the piezoelectric jet unit is lower than the preset pressure threshold, the piezoelectric jet unit corresponding to the jet pressure is determined to be faulty.

10. The obstacle avoidance control method for a transformer oil-immersed internal inspection vehicle according to claim 1, characterized in that, The method for reallocating thrust is as follows: The ratio of the number of failed piezoelectric jet units to the total number of piezoelectric jet units is denoted as the failure ratio. The magnitude of the expected resultant force is reduced and the thrust is redistributed according to the difference between the number 1 and the failure ratio.

Images