Intelligent Control Method and System for Robots Based on Image Analysis Algorithms

Abstract

This invention discloses a robot intelligent control method and system based on image analysis algorithms. The method uses an optimized Single Shot Multi-Box Detector algorithm to identify multi-target features in the working environment image and generate a feature vector set. This feature vector set is then input into the robot's multi-dimensional posture control model to calculate joint angle parameters. These parameters are then combined with the robot's intelligent control parameters to determine a motion planning strategy, controlling the joint drive devices to achieve robot movement. During movement, joint angle and end-effector position information are collected in real time and fed back to the control model. The motion planning strategy is dynamically adjusted through comparative analysis. The system comprises six interconnected units, including image feature recognition and processing, and posture control calculation, each corresponding to a function in each step of the method. This invention achieves full-process intelligent control of the robot, from environmental perception and motion planning to precise control, improving the accuracy and stability of robot operations.

Description

Technical Field

[0001] This invention relates to the field of intelligent robot control, and in particular to a method and system for intelligent robot control based on image analysis algorithms. Background Technology

[0002] With the rapid development of intelligent manufacturing and automation technologies, robots are increasingly being used in industrial production, logistics warehousing, and services. Robot intelligent control technology based on image analysis algorithms has become a key technology for achieving autonomous and precise robot operation. It enables robots to perceive their environment, identify targets, and make autonomous decisions based on environmental changes to complete complex tasks.

[0003] However, existing robot intelligent control technologies have many shortcomings. In terms of target recognition and processing, traditional image analysis algorithms have low efficiency and accuracy in recognizing multiple targets in complex environments, and cannot quickly and accurately locate target objects and feature points. This makes it difficult for robots to accurately perceive the working environment, affecting subsequent control decisions. For example, in scenes with large changes in lighting and complex backgrounds, traditional algorithms are prone to problems such as missed or false detections of targets, preventing robots from performing tasks correctly.

[0004] In terms of motion control strategy formulation and adjustment, existing robot motion planning strategies often lack comprehensive consideration of the robot's own structural parameters, environmental constraints, and real-time feedback information. Traditional motion planning methods cannot adjust in a timely manner according to the robot's actual operating state and environmental changes, which can easily lead to collisions and unstable motion during robot movement. This reduces the robot's work efficiency, poses safety hazards, and fails to meet the operational requirements of high-precision and complex tasks. The robot intelligent control method and system based on image analysis algorithms proposed in this paper effectively solves these problems. Summary of the Invention

[0005] In order to overcome the shortcomings and deficiencies of existing technologies, this invention provides a robot intelligent control method and system based on image analysis algorithms.

[0006] The technical solution adopted in this invention is a robot intelligent control method based on image analysis algorithms, comprising the following steps:

[0007] Step S1: Use the optimized Single Shot Multi-Box Detector algorithm to perform multi-target feature recognition on the acquired robot working environment image data, locate target objects and feature points in the image data, and generate corresponding feature vector sets;

[0008] Step S2: Input the feature vector set into the pre-built robot multi-dimensional posture control model. The model calculates the angle parameters that need to be adjusted for each joint of the robot based on the spatial position information of the target object in the feature vector set and the robot's own initial pose information.

[0009] Step S3: Based on the various parameters of the robot's intelligent control and combined with the joint angle parameters output by the robot's multi-dimensional posture control model, determine the robot's motion planning strategy. The motion planning strategy includes the motion trajectory planning of the robot's end effector and the motion sequence planning of each joint.

[0010] Step S4: Based on the motion planning strategy, control the drive devices of each joint of the robot to rotate each joint according to the predetermined angle parameters and motion sequence;

[0011] Step S5: During the robot's movement, acquire the actual angle information of each joint of the robot and the actual position information of the end effector in real time, and feed them back to the robot's multi-dimensional posture control model.

[0012] Step S6: The robot's multi-dimensional posture control model compares and analyzes the feedback information with the preset target posture information, dynamically adjusts the robot's subsequent motion planning strategy, and performs continuous and precise motion control of the robot.

[0013] Furthermore, the robot's multidimensional posture control model is defined by the following formula:

[0014]

[0015] in, Indicates the robot's first The target adjustment angle of each joint; Weighting coefficients are set according to the robot's structure and operational requirements; Based on the target object feature vector and robot current state feature vector The nonlinear mapping function; For the robot The current angle of each joint; This represents the deviation vector between the current position and the target position of the robot's end effector.

[0016] Furthermore, the various parameters of the robot's intelligent control participate in the construction of the following robot motion planning and decision-making model:

[0017]

[0018] in, This represents the finalized robot motion planning strategy; This is a vector of intelligent control parameters for the robot, including its dynamic parameters, velocity parameters, and acceleration parameters. It is the joint angle parameter vector output by the robot's multidimensional posture control model; This represents a vector containing the robot's operating environment and structural constraints. This is a complex decision function based on the above parameters.

[0019] Furthermore, in step S4, the process of controlling the drive device of each joint of the robot is described by the following formula:

[0020]

[0021] in, This represents the driving force or torque vector output by the drive unit. and For driving control coefficients; It is the target angle vector of each joint output by the robot's multi-dimensional posture control model; This is the feedback vector of the actual angular velocity of each joint of the robot.

[0022] Furthermore, in step S5, the process of feeding back the actual angle information of each joint of the robot and the actual position information of the end effector to the robot's multi-dimensional posture control model is achieved through the following information fusion model:

[0023]

[0024] in, This represents the fused feedback information vector; This is a vector containing the actual angle information of each joint of the robot. It is the vector of actual position information of the robot's end effector; This is the feedback information vector from the previous time step; This is the information fusion function.

[0025] Furthermore, in step S6, the process by which the robot's multi-dimensional posture control model dynamically adjusts the robot's subsequent motion planning strategy is achieved through the following model adjustment:

[0026]

[0027] in, This indicates the adjusted robot motion planning strategy; The original exercise planning strategy; It is the fused feedback information vector; This is a preset target attitude information vector; This is a function for adjusting the motion planning strategy.

[0028] Furthermore, the determination of the robot motion planning strategy in step S3 specifically includes:

[0029] Step S3.1: Based on the joint angle parameters output by the robot's multi-dimensional posture control model, and combined with the speed and acceleration parameters of the robot's intelligent control, calculate the shortest time path for the robot's end effector from its current position to the target position.

[0030] Step S3.2: Based on the range of motion and structural limitations of each joint of the robot, the shortest time path is decomposed into the motion trajectory of each joint;

[0031] Step S3.3: Determine the movement sequence of each joint based on the motion trajectory of each joint and the power parameters of the robot's intelligent control, so as to avoid motion interference and ensure motion stability.

[0032] Furthermore, the drive device for controlling each joint of the robot according to the motion planning strategy in step S4 specifically includes:

[0033] Step S4.1: Activate the corresponding drive devices in sequence according to the motion planning strategy for each joint;

[0034] Step S4.2: According to the motion trajectory of each joint in the motion planning strategy, adjust the output parameters of the drive device in real time so that the joint rotates according to the predetermined angle parameters;

[0035] Step S4.3: During joint rotation, monitor the operating status of the drive device. If any abnormality occurs, immediately activate the emergency control program to adjust the drive device or stop its movement.

[0036] Furthermore, in step S5, the actual angle information of each joint of the robot and the actual position information of the end effector are acquired in real time. The specific operation is as follows:

[0037] Step S5.1: Use angle sensors installed at each joint of the robot to collect the actual rotation angle of the joint in real time, and convert the angle information into digital signals;

[0038] Step S5.2: Obtain the actual position coordinates of the end effector in space in real time using the position sensor installed on the robot's end effector;

[0039] Step S5.3: Encode the acquired joint angle digital signals and end effector position coordinate information to form a feedback information format that can be recognized by the robot's multi-dimensional posture control model.

[0040] A robot intelligent control system based on image analysis algorithms includes:

[0041] The image feature recognition and processing unit is used to perform multi-target feature recognition on the acquired robot working environment image data using the optimized Single Shot Multi-Box Detector algorithm, locate target objects and feature points, and generate corresponding feature vector sets;

[0042] The posture control calculation unit is connected to the image feature recognition and processing unit. It is used to input the feature vector set into the robot's multi-dimensional posture control model and calculate the angle parameters that need to be adjusted for each joint of the robot based on the spatial position information of the target object in the feature vector set and the robot's own initial pose information.

[0043] The motion planning decision unit, connected to the posture control calculation unit, determines the robot's motion planning strategy based on various parameters of the robot's intelligent control and the joint angle parameters output by the robot's multi-dimensional posture control model.

[0044] The joint drive control unit is connected to the motion planning decision unit. Based on the motion planning strategy, it controls the drive devices of each joint of the robot so that each joint of the robot rotates according to the predetermined angle parameters and motion sequence.

[0045] The information acquisition and feedback unit is connected to the joint drive control unit and the attitude control calculation unit. During the robot's movement, it acquires the actual angle information of each joint of the robot and the actual position information of the end effector in real time, and feeds it back to the attitude control calculation unit.

[0046] The motion strategy adjustment unit is connected to the information acquisition and feedback unit and the motion planning and decision-making unit. The robot's multi-dimensional posture control model compares and analyzes the feedback information with the preset target posture information. Through this unit, the robot's subsequent motion planning strategy is dynamically adjusted to achieve continuous and precise motion control.

[0047] Beneficial Effects: This invention proposes a robot intelligent control method and system based on image analysis algorithms. At the target recognition level, an optimized Single Shot Multi-Box Detector algorithm is used to perform multi-target feature recognition on image data of the working environment, accurately locating target objects and feature points and generating feature vector sets. Compared with traditional algorithms, it can better adapt to complex environments, significantly improving the efficiency and accuracy of target recognition, avoiding missed detections and false detections, and laying the foundation for the robot's accurate environmental perception. In terms of motion control, the robot's multi-dimensional posture control model calculates joint angle parameters based on the spatial position of the target object and the robot's initial pose. Simultaneously, it combines various parameters of the robot's intelligent control to construct a motion planning decision model, comprehensively considering the robot's power, speed, acceleration parameters, and environmental and structural constraints to determine the motion planning strategy. Scientific planning is performed from the end effector trajectory to the joint movement sequence. Furthermore, the robot's joint angle and end effector position information are acquired in real time through an information acquisition and feedback unit, processed by an information fusion model, and fed back to the control model. The adjustment model is then used to dynamically optimize the motion planning strategy. Compared with traditional motion control methods, this closed-loop control mechanism can adjust the strategy in real time according to the robot's actual state and environmental changes, effectively avoid collisions, ensure motion stability, significantly improve the robot's work efficiency and operational safety, meet the requirements of high-precision and complex tasks, and achieve continuous and precise motion control of the robot. Attached Figure Description

[0048] Figure 1 This is a flowchart of the method steps of the present invention;

[0049] Figure 2 This is a diagram showing the system unit composition of the present invention. Detailed Implementation

[0050] It should be noted that, unless otherwise specified, the embodiments and features in the embodiments of this application can be combined with each other. The following describes the application in further detail with reference to the accompanying drawings and specific embodiments.

[0051] like Figure 1 As shown, the robot intelligent control method based on image analysis algorithm includes the following steps:

[0052] Step S1: Use the optimized Single Shot Multi-Box Detector algorithm to perform multi-target feature recognition on the acquired robot working environment image data, locate target objects and feature points in the image data, and generate corresponding feature vector sets;

[0053] Specifically, an optimized SSD (Single Shot Multi-Box Detector) algorithm is used to perform multi-target feature recognition on the acquired robot operating environment image data. This involves locating target objects and feature points within the image data and generating corresponding feature vector sets. This step, through the optimized SSD algorithm, employs a multi-scale feature map fusion mechanism and an improved anchor box generation strategy, enhancing the detection accuracy and speed of multiple targets in complex scenes. The algorithm obtains multi-level semantic information from the image through a feature extraction network, combines an attention mechanism to enhance the feature representation of target regions, and uses a non-maximum suppression optimization algorithm to remove redundant detection boxes, ensuring that each real target is uniquely identified. The generated feature vector set contains multi-dimensional information such as the target object's position coordinates, geometric features, and texture features, providing accurate environmental perception data for subsequent multi-dimensional robot posture control models.

[0054] In terms of implementation, this step first preprocesses the input image, including normalization and scaling, to adapt to the optimized input requirements of the SSD algorithm. The algorithm fuses feature maps from different levels through a feature pyramid network to achieve efficient detection of targets of different sizes. Simultaneously, considering the characteristics of robot operation scenarios, the aspect ratio and scale distribution of the anchor frames are optimized, improving the detection capability for irregular targets. In the feature vector generation stage, dimensionality reduction techniques such as principal component analysis are used to compress the feature dimensions while retaining key information, ensuring the efficiency and representativeness of the feature vector set and providing a reliable data foundation for subsequent posture control calculations.

[0055] Step S2: Input the feature vector set into the pre-built robot multi-dimensional posture control model. The model calculates the angle parameters that need to be adjusted for each joint of the robot based on the spatial position information of the target object in the feature vector set and the robot's own initial pose information.

[0056] Specifically, the feature vector set is input into a pre-constructed multi-dimensional robot posture control model. This model calculates the angle parameters that each joint of the robot needs to adjust based on the spatial position information of the target object in the feature vector set and the robot's initial pose information. This model is based on a deep learning architecture, integrating kinematic constraints and dynamic characteristics. Through a multi-input multi-output neural network structure, it achieves a non-linear mapping from the target spatial position to joint angle parameters. During the model training phase, a reinforcement learning strategy is employed, combining forward and inverse kinematics algorithms to optimize network parameters and improve the accuracy of angle calculations. Simultaneously, the model incorporates physical constraints on robot joints and motion smoothness constraints to ensure that the output angle parameters conform to the robot's actual motion capabilities.

[0057] In this step, the spatial position information of the target object in the feature vector set is first converted into a coordinate representation in the robot's coordinate system, and then combined with the robot's current pose information to construct an input tensor. The model processes temporal features through a multilayer perceptron and a long short-term memory network to capture the correlation between the target's motion trajectory and the robot's posture changes. During the angle parameter calculation, a hierarchical optimization strategy is adopted: first, the coarse angles of the main joints are calculated, and then the parameters are refined through a local fine-tuning network to ensure the accuracy of the calculation results. The output angle parameters are kinematically verified to ensure that they meet the robot's kinematic constraints, providing a feasible joint configuration scheme for subsequent motion planning.

[0058] Step S3: Based on the various parameters of the robot's intelligent control and combined with the joint angle parameters output by the robot's multi-dimensional posture control model, determine the robot's motion planning strategy. The motion planning strategy includes the motion trajectory planning of the robot's end effector and the motion sequence planning of each joint.

[0059] Specifically, based on various parameters of the robot's intelligent control and the joint angle parameters output by the robot's multi-dimensional posture control model, a robot motion planning strategy is determined. This strategy includes the trajectory planning of the robot's end effector and the sequence planning of joint movements. This step comprehensively considers factors such as the robot's dynamic parameters, kinematic constraints, and task priorities, and uses a multi-objective optimization algorithm to generate the optimal motion strategy. By constructing a cost function to evaluate the smoothness, energy consumption, and execution time of different trajectory schemes, and combining this with a fast exploration random tree algorithm to search the feasible solution space, the optimal trajectory and joint movement sequence that meet the task requirements are finally determined.

[0060] In the implementation process, a kinematic and dynamic model is first established based on the robot's intelligent control parameters, including constraints such as joint torque limits, speed limits, and acceleration limits. Using the joint angle parameters output by the robot's multi-dimensional posture control model, combined with the predicted motion trend of the target object, multiple candidate trajectory schemes are generated. Each scheme is comprehensively evaluated using a cost function, considering indicators such as trajectory smoothness, obstacle avoidance capability, and energy consumption, to select the optimal trajectory. Simultaneously, a priority scheduling algorithm is used to determine the motion sequence of each joint, ensuring the robot maintains stability and coordination during task execution and avoiding motion interference and singularity problems.

[0061] Step S4: Based on the motion planning strategy, control the drive devices of each joint of the robot to rotate each joint according to the predetermined angle parameters and motion sequence;

[0062] Specifically, based on the motion planning strategy, the drive mechanisms of each robot joint are controlled to enable the robot's joints to rotate according to predetermined angle parameters and motion sequences. This step employs a hierarchical control architecture. The upper-level controller generates the target position sequence for each joint based on the motion planning strategy, while the lower-level controller adjusts the output of the drive mechanism through a closed-loop feedback control algorithm to ensure that the actual motion trajectory of the joints tracks the target trajectory. During the drive control process, an adaptive control algorithm is combined to compensate for model uncertainties and external disturbances, improving control accuracy and system robustness.

[0063] During implementation, the joint angle parameters in the motion planning strategy are first converted into control commands for the drive device, including voltage, current, or pulse signals. A proportional-integral-derivative (PID) controller is used as the basic control algorithm, adjusting the control input in real time based on the deviation between the actual joint angle and the target angle. Simultaneously, feedforward control is introduced to compensate for the system's dynamic characteristics, improving system response speed. During motion, joint angle and velocity information are monitored in real time by sensors. A Kalman filter is used to fuse multi-sensor data, estimate the system state, and predict future trends, providing the controller with more accurate feedback signals to ensure the precision and stability of joint motion.

[0064] Step S5: During the robot's movement, acquire the actual angle information of each joint of the robot and the actual position information of the end effector in real time, and feed them back to the robot's multi-dimensional posture control model.

[0065] Specifically, during robot movement, the actual angle information of each joint and the actual position information of the end effector are acquired in real time and fed back to the robot's multi-dimensional attitude control model. This step collects high-precision motion state data through a sensor network distributed across the robot's joints and end effector. Data synchronization technology is used to ensure the temporal consistency of multi-sensor data, and digital filtering algorithms are used to remove noise interference and extract true and valid motion information. The feedback data undergoes coordinate transformation and format conversion to generate a feedback vector that meets the input requirements of the robot's multi-dimensional attitude control model.

[0066] During implementation, joint angle information is acquired in real time via encoders or resolvers, while end effector position information is obtained through vision sensors or inertial measurement units. A timestamp synchronization mechanism is employed to ensure time alignment of data from different sensors. Kalman filtering or extended Kalman filtering is used to fuse multi-source data, improving the accuracy of state estimation. After preprocessing, the feedback data is transmitted to the robot's multi-dimensional posture control model via a communication interface, providing the model with real-time motion state updates for dynamic adjustment decisions.

[0067] Step S6: The robot's multi-dimensional posture control model compares and analyzes the feedback information with the preset target posture information, dynamically adjusts the robot's subsequent motion planning strategy, and performs continuous and precise motion control of the robot.

[0068] Specifically, the robot's multi-dimensional posture control model compares and analyzes feedback information with preset target posture information, dynamically adjusting the robot's subsequent motion planning strategy to achieve continuous and precise motion control. This step calculates the deviation between the current posture and the target posture using error analysis algorithms, predicts future state change trends using model predictive control technology, and generates adjustment strategies using optimization algorithms. During the adjustment process, system constraints and task priorities are comprehensively considered to ensure both control accuracy and the safety and stability of the robot's motion.

[0069] In implementation, the feedback information and target posture information are first mapped to the same coordinate system, and multi-dimensional deviation indicators such as position error and angle error are calculated. The robot's dynamics model is used to predict state changes at several future moments, and the optimal control input sequence is solved using a rolling optimization algorithm. When adjusting the motion planning strategy, a smooth transition algorithm is employed to avoid abrupt changes and ensure the continuity of robot motion. Simultaneously, the feasibility of the adjusted strategy is verified, checking whether it meets joint limit constraints and environmental limitations. Finally, an executable dynamic adjustment scheme is generated, achieving continuous and precise motion control of the robot in complex environments.

[0070] Preferably, the robot's multidimensional attitude control model has the following formula:

[0071]

[0072] in, Indicates the robot's first The target adjustment angle of each joint; Weighting coefficients are set according to the robot's structure and operational requirements; Based on the target object feature vector and robot current state feature vector The nonlinear mapping function; For the robot The current angle of each joint; This represents the deviation vector between the current position and the target position of the robot's end effector.

[0073] Specifically, the robot's multi-dimensional posture control model achieves accurate prediction and dynamic adjustment of robot joint angles through an innovative parameter fusion mechanism. The weighting coefficients α, β, and γ correspond to the influence of target feature mapping, historical state continuation, and position error compensation, respectively, and are dynamically optimized based on robot structural parameters and operational scenario characteristics using machine learning algorithms. The nonlinear mapping function f comprehensively considers the target object feature vector and the robot's current state vector, employing a deep neural network to model complex relationships in high-dimensional space, ensuring the model's adaptability to environmental changes. By iteratively updating joint angle parameters, the model effectively balances target tracking accuracy and motion smoothness, providing a reliable foundation for robot motion planning and posture control.

[0074] Preferably, the parameters of the robot's intelligent control are used to construct the following robot motion planning decision model:

[0075]

[0076] in, This represents the finalized robot motion planning strategy; This is a vector of intelligent control parameters for the robot, including its dynamic parameters, velocity parameters, and acceleration parameters. It is the joint angle parameter vector output by the robot's multidimensional posture control model; This represents a vector containing the robot's operating environment and structural constraints. This is a complex decision function based on the above parameters.

[0077] Specifically, the robot motion planning decision model systematically integrates the robot's intelligent control parameters, posture adjustment results, and environmental constraints, and uses complex decision functions. Achieving multi-objective optimization. The control parameter vector W contains key indicators such as power, velocity, and acceleration, directly affecting the robot's motion efficiency; the joint angle vector θ provides kinematic constraints; and the environmental constraint vector G incorporates information on workspace limitations and obstacle distribution. Decision function. A hierarchical optimization architecture is adopted, combining genetic algorithms and simulated annealing algorithms to search for the optimal motion strategy in the solution space that satisfies all constraints, ensuring the safety, efficiency and feasibility of robot motion.

[0078] Preferably, in step S4, the process of controlling the drive device for each joint of the robot is described by the following formula:

[0079]

[0080] in, This represents the driving force or torque vector output by the drive unit. and For driving control coefficients; It is the target angle vector of each joint output by the robot's multi-dimensional posture control model; This is the feedback vector of the actual angular velocity of each joint of the robot.

[0081] Specifically, the drive control formula achieves precise control of the robot joint drive mechanism through a two-parameter adjustment mechanism. The drive control coefficients δ and ε adjust the influence weights of the target angle and feedback angular velocity, respectively, and are dynamically tuned based on the robot's dynamics model and load characteristics. The target angle vector provides the position control reference, while the feedback angular velocity vector is used to compensate for system dynamic errors in real time. A Kalman filter algorithm is employed to fuse multi-sensor data to improve feedback accuracy. This control strategy, through a feedforward-feedback composite control architecture, effectively suppresses system nonlinearity and external disturbances, ensuring that the joint motion trajectory tracking error is within the sub-millimeter range.

[0082] Preferably, in step S5, the process of feeding back the actual angle information of each joint of the robot and the actual position information of the end effector to the robot's multi-dimensional posture control model is achieved through the following information fusion model:

[0083]

[0084] in, This represents the fused feedback information vector; This is a vector containing the actual angle information of each joint of the robot. It is the vector of actual position information of the robot's end effector; This is the feedback information vector from the previous time step; This is the information fusion function.

[0085] Specifically, the information fusion model achieves efficient integration and state estimation of multi-source sensor information through a time-series data processing mechanism. It employs a recurrent neural network architecture to process historical feedback information vectors, combining them with current joint angle and position information, and reconstructs the spatiotemporal state using the information fusion function ρ. The model introduces an attention mechanism to automatically assign weights to each sensor's data, adaptively adjusting the fusion strategy for different operational scenarios. Through sliding window filtering and Kalman smoothing techniques, it effectively reduces the impact of measurement noise, outputting continuous and smooth state estimation results, providing a reliable basis for motion strategy adjustment.

[0086] Preferably, in step S6, the process of the robot's multi-dimensional posture control model dynamically adjusting the robot's subsequent motion planning strategy is achieved through the following model adjustment:

[0087]

[0088] in, This indicates the adjusted robot motion planning strategy; The original exercise planning strategy; It is the fused feedback information vector; This is a preset target attitude information vector; This is a function for adjusting the motion planning strategy.

[0089] Specifically, the motion planning adjustment model, based on the principle of model predictive control, achieves dynamic optimization of the robot's motion strategy. This model processes the current motion strategy through a rolling optimization window, integrates feedback information and target posture, and uses a quadratic programming algorithm to solve for the optimal control sequence. The adjustment function σ includes sub-modules such as kinematic feasibility verification, dynamic constraint satisfaction evaluation, and task priority ranking, ensuring that the adjusted strategy conforms to the robot's physical limitations and operational requirements. Through multi-step optimization in the predictive time domain, potential collision risks are avoided in advance, achieving smooth transitions in the motion trajectory and continuity of task execution.

[0090] Preferably, determining the robot motion planning strategy in step S3 specifically includes:

[0091] Step S3.1: Based on the joint angle parameters output by the robot's multi-dimensional posture control model, and combined with the speed and acceleration parameters of the robot's intelligent control, calculate the shortest time path for the robot's end effector from its current position to the target position.

[0092] Step S3.2: Based on the range of motion and structural limitations of each joint of the robot, the shortest time path is decomposed into the motion trajectory of each joint;

[0093] Step S3.3: Determine the movement sequence of each joint based on the motion trajectory of each joint and the power parameters of the robot's intelligent control, so as to avoid motion interference and ensure motion stability.

[0094] Specifically, firstly, based on time-optimal control theory and combined with robot dynamics constraints, the shortest path for the end effector is calculated using the fast travel method. Secondly, the Cartesian space trajectory is mapped to the joint space through inverse kinematics solving, generating reference trajectories for each joint. Finally, based on task decomposition and priority ranking algorithms, the joint movement sequence is determined, and a heuristic search algorithm is used to avoid singular configurations and motion interference. This scheme ensures motion efficiency while reducing energy consumption and extending the robot's lifespan through joint motion trajectory optimization.

[0095] Preferably, the drive device for controlling each joint of the robot according to the motion planning strategy in step S4 specifically includes:

[0096] Step S4.1: Activate the corresponding drive devices in sequence according to the motion planning strategy for each joint;

[0097] Step S4.2: According to the motion trajectory of each joint in the motion planning strategy, adjust the output parameters of the drive device in real time so that the joint rotates according to the predetermined angle parameters;

[0098] Step S4.3: During joint rotation, monitor the operating status of the drive device. If any abnormality occurs, immediately activate the emergency control program to adjust the drive device or stop its movement.

[0099] Specifically, a three-layer cascaded control architecture enables precise control of the drive system. The first layer activates the drive system of the corresponding joint based on a task scheduling algorithm, and uses a state machine model to manage the transitions between different motion stages. The second layer uses an adaptive PID controller to track the reference trajectory in real time and dynamically adjusts control parameters according to load changes. The third layer implements fault diagnosis and fault-tolerant control, monitoring state parameters such as drive current and temperature, and using threshold detection and pattern recognition algorithms to promptly detect anomalies and trigger emergency protection mechanisms. This architecture ensures the reliability and safety of the robot's operation under complex working conditions.

[0100] Preferably, in step S5, the actual angle information of each joint of the robot and the actual position information of the end effector are acquired in real time. The specific operation is as follows:

[0101] Step S5.1: Use angle sensors installed at each joint of the robot to collect the actual rotation angle of the joint in real time, and convert the angle information into digital signals;

[0102] Step S5.2: Obtain the actual position coordinates of the end effector in space in real time using the position sensor installed on the robot's end effector;

[0103] Step S5.3: Encode the acquired joint angle digital signals and end effector position coordinate information to form a feedback information format that can be recognized by the robot's multi-dimensional posture control model.

[0104] Specifically, the data acquisition process utilizes a standardized sensor network to achieve synchronous acquisition and preprocessing of multi-source data. Angle sensors employ absolute encoders to provide high-precision position feedback; position sensors integrate laser rangefinders and visual positioning systems to achieve six-degree-of-freedom pose measurement of the end effector. Data acquisition adopts a distributed architecture, using a fieldbus to achieve time synchronization of sensor nodes. The preprocessing stage implements digital filtering, coordinate transformation, and data compression, employing lossless encoding algorithms to convert raw data into efficient feature vector representations, ensuring the real-time nature and effectiveness of the feedback information.

[0105] like Figure 2 As shown, a robot intelligent control system based on image analysis algorithms is provided. The system includes:

[0106] The image feature recognition and processing unit is used to perform multi-target feature recognition on the acquired robot working environment image data using the optimized Single Shot Multi-Box Detector algorithm, locate target objects and feature points, and generate corresponding feature vector sets;

[0107] The posture control calculation unit is connected to the image feature recognition and processing unit. It is used to input the feature vector set into the robot's multi-dimensional posture control model and calculate the angle parameters that need to be adjusted for each joint of the robot based on the spatial position information of the target object in the feature vector set and the robot's own initial pose information.

[0108] The motion planning decision unit, connected to the posture control calculation unit, determines the robot's motion planning strategy based on various parameters of the robot's intelligent control and the joint angle parameters output by the robot's multi-dimensional posture control model.

[0109] The joint drive control unit is connected to the motion planning decision unit. Based on the motion planning strategy, it controls the drive devices of each joint of the robot so that each joint of the robot rotates according to the predetermined angle parameters and motion sequence.

[0110] The information acquisition and feedback unit is connected to the joint drive control unit and the attitude control calculation unit. During the robot's movement, it acquires the actual angle information of each joint of the robot and the actual position information of the end effector in real time, and feeds it back to the attitude control calculation unit.

[0111] The motion strategy adjustment unit is connected to the information acquisition and feedback unit and the motion planning and decision-making unit. The robot's multi-dimensional posture control model compares and analyzes the feedback information with the preset target posture information. Through this unit, the robot's subsequent motion planning strategy is dynamically adjusted to achieve continuous and precise motion control.

[0112] This paper presents a robot intelligent control method and system based on image analysis algorithms. In terms of target recognition and environmental perception, existing traditional image analysis algorithms struggle to handle complex environments, exhibiting low target recognition efficiency and problems of false positives and false negatives. This method and system utilizes an optimized Single Shot Multi-Box Detector algorithm to perform rapid and accurate multi-target feature recognition on robot operating environment image data. It can not only locate target objects but also capture their feature points and generate feature vector sets. This enables the robot to accurately perceive its surroundings even in scenes with varying lighting and complex backgrounds, providing a reliable basis for subsequent decision-making and fundamentally solving the shortcomings of traditional algorithms in environmental adaptability.

[0113] In the motion planning and control stage, existing technologies lack comprehensive consideration of robot parameters, environmental constraints, and real-time feedback, leading to unreasonable motion planning and unstable execution. This system utilizes a multi-dimensional robot posture control model, combining the target object's spatial position and the robot's initial pose to calculate joint angle parameters. Then, it leverages the robot's intelligent control parameters such as power and speed to construct a motion planning decision model, scientifically planning the end effector's trajectory and joint movement sequence. Simultaneously, during motion, the information acquisition and feedback unit acquires joint angle and end effector position information in real time, which is then fused and fed back to the control model. The motion strategy adjustment unit then dynamically optimizes the motion planning. This multi-model collaborative, real-time feedback and adjustment mechanism allows the robot to avoid collisions and ensure stable motion, significantly improving work efficiency and safety, and is significantly superior to traditional motion control technologies.

[0114] At the system architecture level, the six main modules of this intelligent control system have clearly defined roles and work closely together. The image feature recognition and processing unit, posture control and calculation unit, motion planning and decision-making unit, etc., cooperate sequentially, forming an efficient and orderly control process from environmental information acquisition to motion strategy generation, and then to drive control and feedback adjustment. Compared with the loose architecture and poor coordination of traditional systems, this system, through modular design and tight connection, achieves precise and efficient intelligent control of the robot, providing solid technical support for the application of robots in complex task scenarios.

[0115] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "set," "install," "connect," "link," and "fix" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal communication between two components. Those skilled in the art will understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0116] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various equivalent changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A robot intelligent control method based on image analysis algorithms, characterized in that, Includes the following steps: Step S1: Use the optimized Single Shot Multi-Box Detector algorithm to perform multi-target feature recognition on the acquired robot working environment image data, locate target objects and feature points in the image data, and generate corresponding feature vector sets; Step S2: Input the feature vector set into the pre-built robot multi-dimensional posture control model. The model calculates the angle parameters that need to be adjusted for each joint of the robot based on the spatial position information of the target object in the feature vector set and the robot's own initial pose information. Step S3: Based on the various parameters of the robot's intelligent control and the joint angle parameters output by the robot's multi-dimensional posture control model, determine the robot's motion planning strategy. The motion planning strategy includes the motion trajectory planning of the robot's end effector and the motion sequence planning of each joint. Step S4: Based on the motion planning strategy, control the drive devices of each joint of the robot to rotate each joint according to the predetermined angle parameters and motion sequence; Step S5: During the robot's movement, acquire the actual angle information of each joint of the robot and the actual position information of the end effector in real time, and feed them back to the robot's multi-dimensional posture control model. Step S6: The robot's multi-dimensional posture control model compares and analyzes the feedback information with the preset target posture information, dynamically adjusts the robot's subsequent motion planning strategy, and performs continuous motion control of the robot. The robot's multidimensional attitude control model is defined by the following formula: ； in, Indicates the robot's first The target adjustment angle of each joint; Weighting coefficients are set according to the robot's structure and operational requirements; Based on the target object feature vector and robot current state feature vector The nonlinear mapping function; For the robot The current angle of each joint; This represents the deviation vector between the current position and the target position of the robot's end effector; The parameters of the robot's intelligent control are used to construct the following robot motion planning and decision-making model: ； in, This represents the finalized robot motion planning strategy; This is a vector of intelligent control parameters for the robot, including its dynamic parameters, velocity parameters, and acceleration parameters. It is the joint angle parameter vector output by the robot's multidimensional posture control model; This represents a vector containing the robot's operating environment and structural constraints. This is a complex decision function based on the above parameters.

2. The robot intelligent control method based on image analysis algorithm according to claim 1, characterized in that, Step S4, the process of controlling the drive devices of each joint of the robot, is expressed by the following formula: ； in, This represents the driving force or torque vector output by the drive unit. and For driving control coefficients; It is the target angle vector of each joint output by the robot's multi-dimensional posture control model; This is the feedback vector of the actual angular velocity of each joint of the robot.

3. The robot intelligent control method based on image analysis algorithm according to claim 1, characterized in that, Step S5, which involves feeding back the actual angle information of each joint of the robot and the actual position information of the end effector to the robot's multi-dimensional attitude control model, is achieved through the following information fusion model: ； in, This represents the fused feedback information vector; This is a vector containing the actual angle information of each joint of the robot. It is the vector of actual position information of the robot's end effector; This is the feedback information vector from the previous time step; This is the information fusion function.

4. The robot intelligent control method based on image analysis algorithm according to claim 1, characterized in that, Step S6, the process by which the robot's multi-dimensional posture control model dynamically adjusts the robot's subsequent motion planning strategy, is achieved through the following model adjustment: ； in, This indicates the adjusted robot motion planning strategy; The original exercise planning strategy; It is the fused feedback information vector; This is a preset target attitude information vector; This is a function for adjusting the motion planning strategy.

5. The robot intelligent control method based on image analysis algorithm according to claim 1, characterized in that, Step S3, determining the robot motion planning strategy, specifically includes: Step S3.1: Based on the joint angle parameters output by the robot's multi-dimensional posture control model, and combined with the robot's intelligent control speed and acceleration parameters, calculate the shortest time path from the robot's end effector to the target position. Step S3.2: Based on the range of motion and structural limitations of each joint of the robot, the shortest time path is decomposed into the motion trajectory of each joint; Step S3.3: Determine the movement sequence of each joint based on the motion trajectory of each joint and the power parameters of the robot's intelligent control, so as to avoid motion interference and ensure motion stability.

6. The robot intelligent control method based on image analysis algorithm according to claim 1, characterized in that, Step S4, which controls the drive devices of each joint of the robot according to the motion planning strategy, specifically includes: Step S4.1: Activate the corresponding drive devices in sequence according to the motion planning strategy for each joint; Step S4.2: According to the motion trajectory of each joint in the motion planning strategy, adjust the output parameters of the drive device in real time so that the joint rotates according to the predetermined angle parameters; Step S4.3: During joint rotation, monitor the operating status of the drive device. If any abnormality occurs, immediately activate the emergency control program to adjust the drive device or stop its movement.

7. The robot intelligent control method based on image analysis algorithm according to claim 1, characterized in that, Step S5, which involves acquiring the actual angle information of each joint of the robot and the actual position information of the end effector in real time, specifically includes: Step S5.1: Use angle sensors installed at each joint of the robot to collect the actual rotation angle of the joint in real time, and convert the angle information into digital signals; Step S5.2: Obtain the actual position coordinates of the end effector in space in real time using the position sensor installed on the robot's end effector; Step S5.3: Encode the acquired joint angle digital signals and end effector position coordinate information to form a feedback information format for the robot's multi-dimensional posture control model to recognize.

8. A robot intelligent control system based on image analysis algorithms, characterized in that, This system is applied to the robot intelligent control method based on image analysis algorithm as described in claim 1, comprising: The image feature recognition and processing unit is used to perform multi-target feature recognition on the acquired robot working environment image data using the optimized Single Shot Multi-Box Detector algorithm, locate target objects and feature points, and generate corresponding feature vector sets; The posture control calculation unit is connected to the image feature recognition and processing unit. It is used to input the feature vector set into the robot's multi-dimensional posture control model and calculate the angle parameters that need to be adjusted for each joint of the robot based on the spatial position information of the target object in the feature vector set and the robot's own initial pose information. The motion planning decision unit, connected to the posture control calculation unit, determines the robot's motion planning strategy based on various parameters of the robot's intelligent control and the joint angle parameters output by the robot's multi-dimensional posture control model. The joint drive control unit is connected to the motion planning decision unit. Based on the motion planning strategy, it controls the drive devices of each joint of the robot so that each joint of the robot rotates according to the predetermined angle parameters and motion sequence. The information acquisition and feedback unit is connected to the joint drive control unit and the attitude control calculation unit. During the robot's movement, it acquires the actual angle information of each joint of the robot and the actual position information of the end effector in real time, and feeds it back to the attitude control calculation unit. The motion strategy adjustment unit is connected to the information acquisition and feedback unit and the motion planning and decision-making unit. The robot's multi-dimensional posture control model compares and analyzes the feedback information with the preset target posture information. The motion strategy adjustment unit dynamically adjusts the robot's subsequent motion planning strategy to achieve continuous and precise motion control of the robot.

Images