14results about How to "Fast tracking" patented technology

The invention belongs to the field of deep reinforcement learning and automatic control, and particularly relates to a four-rotor unmanned aerial vehicle hanging anti-swing control method based on PPO-Clip. The method comprises the following steps: 1) establishing a mathematical model of a suspended unmanned aerial vehicle system based on Newton-Euler and Euler-Lagrange methods; 2) according to the system model, designing a Backstepping-PID controller to carry out pose control on the unmanned aerial vehicle; and 3) constructing a deep reinforcement learning framework based on a near-end strategy optimization algorithm (PPO-Clip), wherein the deep reinforcement learning framework comprises the processes of sampling of an unmanned aerial vehicle experience sequence, reward function design, intelligent agent training and the like. According to the invention, by designing the four-rotor unmanned aerial vehicle suspension anti-swing control system based on the PPO-Clip algorithm, a reasonable reward function is designed to guide the unmanned aerial vehicle to adjust the position and attitude of the unmanned aerial vehicle, so that the suspension unmanned aerial vehicle can track an expected trajectory with high precision under the condition of ensuring that the load stops or converges in a short time; therefore, the trajectory tracking performance, the load anti-swing performance, the anti-interference performance and the like of the four-rotor hanging unmanned aerial vehicle are improved.

The present invention discloses a method, device and storage medium for detecting and tracking birds in a substation based on deep learning. The method comprises: obtaining real-time monitoring video images near the substation; using a deep learning method to identify birds in each video image; performing multi-target tracking on all bird entities identified in the video image; performing single-target tracking on each bird in the video image; calculating the safe distance of each bird according to the coordinate interval of each bird on the most recent frame of the video image and the coordinate interval of the nearest substation equipment on the most recent frame of the video image; if the future flight distance of the bird on the image video exceeds the safe distance of the bird, the current bird is determined to be an expulsion target bird, and the video image with the expulsion target bird marked is sent to an alarm device, and an expulsion instruction is sent to an expulsion device; when the portion of the area occupied before and after the expulsion is reduced exceeds a threshold, the expulsion is determined to be successful. The present invention has high tracking accuracy and tracking speed.

The present invention discloses a twin network target tracking algorithm with enhanced positioning quality information, including: determining the target, obtaining the first frame image, determining the annotation box using known information, cropping the target area, and obtaining the target template; similarity comparison, obtaining the current frame, using the feature extraction subnetwork to obtain the feature map of the target and the search area and perform similarity comparison to obtain a similarity matrix; target positioning, the fusion classification branch performs the classification task based on the similarity matrix, and outputs the similarity prediction score; boundary prediction, the regression branch predicts the target boundary based on the similarity matrix, and selects the best prediction box in combination with the similarity prediction score; outputs the tracking result, outputs the target boundary box, and performs target tracking for the next frame. The present invention can improve the classification and boundary prediction process, introduces the CNN network to make the model more fully utilize the semantic information in the feature map, and the prediction result has a higher credibility.

The civil aircraft demand analysis method based on the knowledge graph utilizes the characteristics of the knowledge graph and the graph database in explainability and high efficiency in multi-hop relationship search, finally improves the speed of demand link completeness calculation and demand link tracking; utilizes the characteristics that the knowledge graph can understand the demand sentence text features, finally can realize the recommendation of the demand link and the calculation of the demand design completeness.

The present application relates to radar detection technology field, especially to a kind of signal processing method of Doppler frequency offset interference of height-finding radar, the method is by monitoring the maneuverability of height-finding radar and dynamically adjusting frequency modulation period and frequency modulation bandwidth of symmetrical frequency modulation signal;Based on the frequency modulation period and frequency modulation bandwidth after adjustment generate positive slope frequency modulation signal and negative slope frequency modulation signal and carry out periodic alternate emission, collect and process and difference calculation first echo signal and second echo signal after reflection, obtain after eliminating Doppler frequency offset interference ranging frequency difference, further avoid the height-finding error caused by Doppler shift amount, to obtain accurate target height based on ranging frequency difference calculation.This application is by using the Doppler shift polarity opposite characteristics of positive and negative slope frequency modulation signal, by frequency spectrum line pairing and difference calculation offset Doppler shift component, finally based on the target height of corrected frequency information inversion, to realize high-precision height-finding.

The present application relates to a distributed photovoltaic MPPT method, system, device and medium, and relates to the technical field of distributed photovoltaic power generation, which comprises the following steps: preprocessing photovoltaic array data, obtaining irradiance, temperature and MPP voltage dataset; determining the optimal ElmanNN hidden layer node number by K-fold cross-validation and RMSE evaluation, and dividing the dataset into a training set, a validation set and a test set; inputting the training set into global search of the firefly algorithm and local optimization of the ant colony optimization algorithm, defining the stable solution as the FA-EAS optimal parameters after convergence; updating the ElmanNN weight threshold based on the optimal parameters, deploying the training model to perform MPP voltage online prediction; inputting the difference between the predicted voltage and the reference voltage into a PID controller, calculating the control amount through the incremental PID, generating a PWM signal to control the duty cycle of the Boost converter, and realizing maximum power point tracking. The present application can improve the tracking speed, stability and success rate of MPPT.

The application discloses a deep spatio-temporal correlation video target tracking method and system, which realizes target tracking for each video sequence and guarantees to obtain accurate video target tracking results. Firstly, a spatio-temporal feature extractor is designed to extract spatio-temporal features of a template sequence and a search sequence. Secondly, a feature matching module is introduced, which is composed of a classification branch and a regression branch. The extracted template spatio-temporal features and the search spatio-temporal features are subjected to similarity matching through correlation filtering to obtain multi-channel correlation filtering features respectively. Then, a target tracking module is deployed, which includes a classification head and a regression head, and the classification score map and the regression score map are calculated according to the input multi-channel correlation filtering features respectively, which are used to predict the target position and estimate the target scale. Finally, the spatio-temporal correlation visual tracking model is optimized by minimizing the defined joint loss. In the target tracking test, a confidence region estimation strategy is proposed, so that the target tracking is kept robust and accurate in the video sequence.

This invention relates to the field of adaptive control technology, specifically disclosing a tool wear adaptive compensation system based on multi-dimensional sensor data fusion. It acquires force, vibration, and acoustic emission signals in real time to obtain a fused feature sequence; extracts the dynamic force gradient and calculates the fractal dimension, decomposing it into long-trend and short-period components to generate a coupled modulation sequence, establishing a nonlinear mapping with two latent states: tool wear amount and workpiece hardness; employs a dual-state unscented Kalman filter to jointly estimate wear amount and hardness, using a cross-covariance matrix to decouple their contributions to the sensor signals; calculates dimensional deviations based on wear estimates, and uses hardness estimates to eliminate hardness fluctuation interference, forming a compensation amount specific to wear, which is then written into the CNC system. Simultaneously, it monitors signal changes after compensation to generate confidence levels, and uses feedback to adjust the filter covariance for iterative optimization. This invention can separate the interference of material hardness fluctuations on wear estimation, avoid erroneous compensation, and improve batch processing accuracy.

The invention relates to a correlation filtering method for enhancing the tracking performance of an unmanned aerial vehicle, and the method comprises the steps: S1, carrying out the collection of an original image through an RGB camera, and training a tracking model of a discriminant unmanned aerial vehicle; s2, performing feature extraction on the new image by fusing CN and fHOG features to obtain new sample information; s3, performing dimensionality reduction processing on the obtained features by using a principal component analysis (PCA) mode to obtain sample information with smaller dimensionality and lower complexity so as to improve the tracking speed; s4, performing similarity comparison on the obtained sample information and the trained model information by using an elastic regularization network, judging whether learning is performed or not, discarding image information with relatively low similarity, and determining that the learned image information is trained to become a new tracking model; s5, optimizing the trained tracker model by using an alternating direction multiplier method (ADMM); and S6, target position information is determined, and unmanned aerial vehicle tracking is realized through continuous iteration. The tracking rate and the accuracy rate in the method are both at a high level.

This application relates to target tracking and provides a multi-target tracking method. It acquires measurement data of the target to be tracked using at least two sensors to obtain a first set of measurement data pairs. Then, based on preset constraints, related measurement data pairs in this first set are eliminated to obtain the true target. Next, the starting point of the true target's trajectory is determined using an FCM algorithm and the measurement data pairs of the true target. Then, the true target is correlated using the FCM algorithm and the real-time measurement data of the true target to obtain a stable trajectory. Finally, the true target is tracked based on the starting point and the stable trajectory. This application eliminates false targets using preset constraints and obtains the starting point of the trajectory based on the FCM algorithm in the trajectory initiation part, and uses FCM to determine the stable trajectory of the true target in the trajectory correlation part, thereby achieving the tracking of the true target with short processing time and fast tracking speed.

This invention discloses a method for integrating frequency detection function into a software phase-locked loop (PLL) of a three-phase converter, relating to the field of data processing. The method includes the following steps: Step S1, obtaining the grid voltage in a three-phase stationary coordinate system and converting it into a coordinate system grid voltage. Compared with existing technologies, the advantages of this invention are: This invention overcomes the limitations of low-bandwidth PLLs on frequency response, significantly improving the dynamic response speed and steady-state accuracy of frequency detection, while also possessing good anti-interference capabilities, solving the problem of balancing response speed and detection accuracy in traditional solutions; without changing the PLL PI parameters to ensure converter impedance stability and weak grid adaptability, the calculated real-time angular frequency is used as a feedforward value to replace the fixed feedforward, effectively improving the PLL's tracking speed for grid frequency changes; and it saves software resource costs.

The present invention provides an implementation method and control system of an anti-disturbance control based on expected dynamics, wherein the control implementation method includes the following steps: describing the controlled industrial system using a first-order inertia plus pure delay system; designing the expected dynamic equation of the closed-loop system; based on the designed expected dynamic equation, calculating the output value of the controlled industrial system and the control quantity of the closed-loop system through an extended state observer algorithm, obtaining the tracking value of the output value of the next calculation step of the controlled industrial system and the tracking value of the next calculation step of the total disturbance of the system; based on the obtained tracking value and the expected dynamic equation, obtaining the input values ​​of the next two calculation steps of the controlled industrial system through control law calculation, and the closed-loop system adjusts the change amount of the actuator in real time through the calculation results. The present invention can significantly reduce the difficulty of parameter setting in the implementation of anti-disturbance control, quantify the parameters in the implementation of anti-disturbance control, and has a good control effect.

This invention relates to the field of adaptive control technology, specifically disclosing a tool wear adaptive compensation system based on multi-dimensional sensor data fusion. It acquires force, vibration, and acoustic emission signals in real time to obtain a fused feature sequence; extracts the dynamic force gradient and calculates the fractal dimension, decomposing it into long-trend and short-period components to generate a coupled modulation sequence, establishing a nonlinear mapping with two latent states: tool wear amount and workpiece hardness; employs a dual-state unscented Kalman filter to jointly estimate wear amount and hardness, using a cross-covariance matrix to decouple their contributions to the sensor signals; calculates dimensional deviations based on wear estimates, and uses hardness estimates to eliminate hardness fluctuation interference, forming a compensation amount specific to wear, which is then written into the CNC system. Simultaneously, it monitors signal changes after compensation to generate confidence levels, and uses feedback to adjust the filter covariance for iterative optimization. This invention can separate the interference of material hardness fluctuations on wear estimation, avoid erroneous compensation, and improve batch processing accuracy.

The invention discloses an energy regulation method for a photovoltaic system, and relates to the field of energy regulation for a photovoltaic system. The technical scheme thereof mainly includes: collecting comprehensive state data within a time period T, and processing the comprehensive state data to obtain time series state data; using a sliding average filter to preliminarily process the time series state data to obtain a smoothed initial data set, using the initial data set as an observation input of a Kalman filter, further smoothing the data using the Kalman filter to obtain a high-precision gain data set, and using an improved P&O algorithm to perform disturbance judgment on the gain data set; using an IncCond method to eliminate power oscillations of a voltage in a prediction interval, and performing segmented demand judgment, using a segmented search method to accurately locate a maximum power point, and outputting it as a system operating point, thereby realizing energy regulation of the photovoltaic system at the maximum power point and improving the energy efficiency of the photovoltaic system.