346 results about "Algorithm complexity" patented technology

The invention discloses a ship block construction process-level production scheduling method and system based on a particle swarm-genetic hybrid algorithm. The method comprises the following steps: constructing a standardized process-level process design data model; based on an operation plan decomposition rule, decomposing the segmented production task into a part-level operation production task matched with the process; establishing an optimization target model with the shortest delivery time, the least overdue tasks, the most balanced load and the like, and constraint rules such as lane and line separation and customization station rules, process route and process constraint rules, material and capacity constraint, priority and the like; optimizing the segmented production plan through a particle swarm-genetic hybrid algorithm until an optimization result meeting the requirement is obtained; and according to the feedback of the actual execution condition of the plan, the orders are classified based on a clustering algorithm, and the operation cycle parameters of different processes are dynamically compensated. The method has the advantages of being high in plan fineness, small in operation progress estimation deviation and the like, is high in calculation speed and low in algorithm complexity, fully considers support of process design data of a design end on algorithm implementation, and is very suitable for engineering application.

The embodiment of the invention provides a data management method and device, computer equipment and a storage medium. The data management method comprises the following steps: obtaining audio and video data to be transmitted and a sensitivity level corresponding to the audio and video data; at least according to the sensitivity level corresponding to the audio and video data, a matched transmission strategy is configured for the audio and video data, and the matched transmission strategy comprises an encrypted transmission strategy or a non-encrypted transmission strategy; the parameters of the encryption transmission strategy comprise at least one of the encryption algorithm type, the encryption algorithm complexity, the encryption key length and the encryption key complexity; and transmitting the audio and video data based on the matched transmission strategy.

The application discloses a kind of based on 3D printing's three-dimensional model slice path processing method and device. Including: control 3D printer to starting point as starting point;Based on the region surrounded between starting point and reference point, determine the printing point to be printed that 3D printer needs to execute printing action during executing 3D printing job;Starting point and the printing point to be printed are sequentially connected by smooth straight line, generate printing path;Based on printing path control 3D printer executes 3D printing job and obtains target object, to effectively reduce the number of line segments of slice contour line of three-dimensional model forming target object, avoid 3D printer in printing job, in the pause frequency of line segment head, tail, simultaneously avoid the problem that 3D printing model surface protrusion is caused by long time scanning due to pause.In reducing pause frequency, printing is more smooth, improve printing efficiency, simultaneously also reduce algorithm complexity, help the simplification of algorithm.And effectively simplify the algorithm of 3D printer.

The invention discloses a hyperspectral image classification method and system based on a double-branch lightweight algorithm, belongs to the technical field of computer vision and image processing, and solves the problems that an existing image classification model is complex in structure, so that the calculated amount is large, the occupied memory is high, and balance between classification precision and model lightweight is difficult to achieve. The method comprises the steps of collecting a hyperspectral image, preprocessing the hyperspectral image, performing iterative training on a double-branch lightweight algorithm model based on a training set, executing the double-branch lightweight algorithm model, and classifying a test set by the double-branch lightweight algorithm model. According to the hyperspectral image classification method, the algorithm complexity is greatly reduced, the high precision of hyperspectral image classification is ensured, the lightweight space-spectrum Transformer is introduced into the double-branch lightweight algorithm model, and through the optimization design of the whole structure and the lightweight efficient Attention-Aware mechanism, the classification accuracy of the hyperspectral image is improved. The complexity of the model is reduced, and the characteristic relation between the modeling space and the spectrum sequence can be better established.

The invention relates to the technical field of the Internet of Things, in particular to an edge computing privacy task unloading method, system and device based on reinforcement learning, and the method comprises the steps: obtaining computing task information and task unloading selection generated by each terminal device, and the task unloading selection comprises task selection local execution or task selection unloading to an edge server, local execution cost or unloading cost of each task is obtained based on the calculation; performing statistics to obtain the total cost of all tasks, and establishing an initial global optimization target; the privacy entropy of the task is calculated, and the privacy entropy of the task is used as unloading additional cost to be embedded into the initial global optimization target to update the global optimization target; and searching an optimal strategy of the updated global optimization target by adopting a single-agent reinforcement learning algorithm. The privacy guarantee strength and the dynamic scene adaptability are improved while the algorithm complexity is greatly reduced and the algorithm efficiency is improved, and the method can be suitable for large-scale node, high-dynamic and strong-privacy-constraint application scenes in edge calculation.

The invention discloses a GNSS (Global Navigation Satellite System) signal spoofing detection method, device and equipment, a storage medium and a program product. The method comprises the following steps: tracking and monitoring a plurality of GNSS satellite signals, calling a tracking loop correlator to carry out autocorrelation operation to obtain the autocorrelation power of each satellite signal, constructing the AP detection quantity of each satellite signal based on the autocorrelation power, verifying the Gaussian property of the AP detection quantity, and carrying out moving average filtering processing on the AP detection quantity conforming to Gaussian distribution to obtain the AP detection quantity of each satellite signal. The method comprises the following steps: acquiring Gaussian distribution characteristic information of a GNSS satellite signal, constructing an AP-MA detection quantity, determining a detection threshold according to the Gaussian distribution characteristic information, detecting the AP-MA detection quantity based on the detection threshold, and judging whether the GNSS satellite signal has deception jamming or not; the AP detection quantity is constructed based on the self-correlation power to carry out signal spoofing detection, so that the detection range and the detection precision are effectively improved, and the timeliness and the stability of spoofing detection are ensured under the condition of not increasing additional equipment and algorithm complexity.

The invention provides a TOF (Time of Flight) estimation method of a matrix bundle based on partitioning and random SVD (Singular Value Decomposition). The method comprises the following steps: firstly, acquiring channel state information (CSI) by using a commercial WiFi device, and constructing a CSI matrix; and secondly, performing block overlapping processing on the CSI matrix, constructing an augmented matrix for each CSI data block, and performing SVD decomposition on the augmented matrix, thereby reducing the calculation complexity of SVD decomposition. And then, the time of flight (TOF) of each CSI data block is estimated by using a matrix pencil algorithm, and the minimum TOF is direct path estimation. According to the method designed by the invention, the robustness of the system and the estimation precision of the TOF are improved under the condition that the algorithm complexity is greatly reduced.

The invention provides a multi-face cutting method and device for a three-dimensional grid, electronic equipment and a storage medium. According to the embodiment of the invention, a cutting face set of the three-dimensional grid is called to calculate a cutting face surrounding frame; obtaining a small grid to be cut through the K-D tree and / or obtaining a large grid to be cut through the octree of the self-adaptive voxel; then decomposing a small grid to be cut and a large grid to be cut into at least one polygonal surface patch, constructing vertexes of the polygonal surface patch into an effective point set to be processed according to a clockwise sequence, and retaining, eliminating and / or newly adding effective points according to a relative position relation to construct a three-dimensional surface patch; and finally, merging the three-dimensional surface patches into the three-dimensional grid to form a new three-dimensional grid. By adopting the scheme, surface patches can be quickly positioned and screened, and the performance and efficiency are balanced; the points and the edges are regenerated in an ordered point mode, the calculation efficiency is improved, the algorithm complexity is reduced, and the effect of efficiently and accurately performing multi-face cutting on the city-level three-dimensional grid is achieved.

The invention belongs to the field of analog integrated circuit design, and particularly relates to a miniaturized high-precision non-binary R-2R DAC with digital correction. Non-binary characteristics are introduced by setting resistance values of bridge resistors RBm and RBm + 1, so that an LSB section is a binary weight section, and ISB and MSB are non-binary weight sections; therefore, the width-to-length ratios of the MOS tubes of the switches SWm + 1-SWm + i + j and SW1 are equal and do not need to be increased proportionally, the non-binary characteristic is further introduced by setting the switch on-resistance, the selection of the unit resistance R only needs to meet the matching precision of the LSB section, and the area of the switch resistor array is greatly reduced. Static errors in the ISB segment and the MSB segment are respectively corrected by using an OEM correction technology, only digital code redundancy between adjacent segments of the LSB segment, the ISB segment and the MSB segment needs to be obtained through measurement, redundant digital codes are removed by adjusting digital input codes of corresponding segments, and the m + n + k-1 bit binary DAC is obtained through calibration, so that the memory of a digital calibration algorithm is efficient, and the algorithm complexity is effectively reduced. According to the invention, the area of the DAC is obviously reduced while high precision is realized.

The invention relates to an optimal consistency control method for a random multi-agent system with unknown control gain. The optimal consistency control method is suitable for leader following consistency control of the random multi-agent system with unknown control gain. According to the method, the system uncertainty is approximated by constructing a self-adaptive identifier; designing an observer based on a neural network to reconstruct an unknown control gain; a dynamic event triggering mechanism is introduced, an auxiliary dynamic variable and a dynamic threshold parameter are used for reducing the updating frequency of a controller, and communication and energy consumption are reduced; in combination with an act-critic reinforcement learning framework and based on a positive function related to a Bellman residual error, a neural network weight update rate is constructed, an optimal consistency controller is designed, and the algorithm complexity is simplified while the system stability is ensured. The method has high robustness and is suitable for efficient cooperative control of a multi-agent system in a complex environment.

The invention provides a wavelet-CNN satellite communication interference signal detection method and system based on data driving, and belongs to the field of satellite communication signal processing. The problems that an existing signal detection technology is high in algorithm complexity and low in interference signal detection accuracy are solved. According to the method, the ground radiation source communication signal received by the satellite is de-noised by fusing the soft and hard threshold functions to improve the wavelet threshold function, so that noise and interference can be effectively suppressed, and meanwhile, key detail information of the signal is reserved; two-dimensional time-frequency data are constructed through Fourier transform and normalized, a convolutional neural network is trained to enhance the expression ability of signal features, and finally a signal detection result is tested and output. According to the method, the reliability of signal processing can be improved, the overall performance of a satellite-borne communication system can be optimized, and an effective solution is provided for signal monitoring in a complex channel and a dynamic environment.

The invention provides an inter-bank loan fund risk prompting method and device, equipment and a storage medium, and relates to the field of privacy computing, and the method comprises the steps: obtaining the first account information of a first account and the transfer information corresponding to the loan fund when it is detected that the loan fund in the first account is transferred to a second account of a second bank; performing encryption processing on the first account information to obtain a corresponding ciphertext; sending the ciphertext and the transfer information to a second bank, so that the second bank determines corresponding second account information based on the transfer information in a trusted execution environment, and performs loan fund risk query; and receiving a query result fed back by the second bank, and generating a loan fund risk prompt message under the condition that the query result indicates that the risk exists. According to the method, sensitive data are guaranteed not to be leaked by utilizing a hardware isolation mechanism of the trusted execution environment, and meanwhile, plaintext calculation is realized in the trusted execution environment, so that the algorithm complexity is remarkably reduced, and the timeliness of loan fund tracking is improved.

The invention belongs to the technical field of digital image processing, and particularly relates to a regional hole filling method based on run-length coding. According to the method, the candidate background area is directly obtained by negation of the original area and construction of the extended rectangular area, the complexity of boundary processing in traditional hole filling is simplified, the area is segmented by utilizing run length coding, and then dual adjacency relation judgment is combined, so that the accuracy of hole filling is improved. The connectivity among the codes in the background area can be identified more accurately, so that the accuracy of extraction of the connected area is improved, the mark numbers of the connected run-length codes are recorded in a mask mode, so that the run-length codes do not need to be additionally reordered when the background area is subsequently summarized, the computing resources are saved, the algorithm complexity is reduced, and the efficiency is improved. The integrity of the candidate background area is guaranteed through the extended rectangles, accurate extraction of the background area is guaranteed through depth-first search, the method can adapt to filling of original area holes of various shapes and sizes, and good robustness is achieved.

The invention discloses an aerospace network calculation unloading and trajectory collaborative optimization method based on reinforcement learning. In order to solve the problem of dynamic task response hysteresis caused by satellite high-altitude delay, limited unmanned aerial vehicle coverage and resource dispersion, an original problem is decoupled into two sub-problems of task unloading proportion distribution and unmanned aerial vehicle trajectory planning by constructing a mixed integer nonlinear programming model and taking minimization of system task average time delay as a target. According to the method, a K-means clustering algorithm is adjusted by utilizing attributes, dynamic task distribution characteristics are quantified, and track reference points are generated to simplify a state space; and meanwhile, a QPLEX-based multi-agent reinforcement learning framework is designed, and collaborative decision-making of the satellite and the unmanned aerial vehicle is realized. According to the method, a time delay perception optimization mechanism is established by fusing the trajectory reference points generated by clustering, closed-loop joint optimization of an unloading decision and a flight trajectory is formed, the algorithm complexity and the communication overhead are remarkably reduced, and the aerospace network resource scheduling efficiency is effectively improved.

The invention discloses a three-dimensional millimeter wave time-varying channel estimation method and device, and belongs to the field of wireless communication, and the method comprises the steps: building a millimeter wave communication system containing array motion; a three-dimensional time-varying channel model considering the Doppler effect is established, composite signals, caused by array motion, in a three-dimensional time-varying channel are analyzed, each column of a direction matrix after synthesis is consistent with the array manifold of a planar array, and it is indicated that a moving linear array can form a virtual planar array; sparse representation of a three-dimensional time-varying channel is realized through linear array motion; a three-dimensional time-varying channel is recovered based on an iterative meshless weighting algorithm of array motion; comprising AoAs estimation and path gain estimation. According to the method, the negative influence of Doppler frequency shift in the time-varying channel on channel estimation is overcome, the virtual array element is constructed, channel estimation of the three-dimensional millimeter wave time-varying channel is realized through linear array motion, the channel estimation precision is improved, and the pilot frequency overhead and the algorithm complexity are greatly reduced.

The invention discloses a ship 3D target detection method and system based on image and laser radar fusion in the field of ship navigation intellectualization, and the method comprises the steps: obtaining a two-dimensional bounding box set of an object in a camera image through a Yolov11 detection frame, converting the eight vertex sets of a three-dimensional bounding box into a camera coordinate system, obtaining the minimum bounding rectangle coordinate of a plane coordinate set, and obtaining the minimum bounding rectangle coordinate of the plane coordinate set; calculating an IOU value for each minimum enclosing rectangle and a two-dimensional bounding box in the two-dimensional bounding box set, projecting a plane coordinate of the three-dimensional bounding box before the minimum enclosing rectangle is converted to a two-dimensional plane, and obtaining a three-dimensional detection frame and category information; information such as the position and the size of an object existing in the navigation direction is obtained in real time through a laser radar and fused with a camera picture; the improved DBSCAN clustering algorithm is designed, the core block is confirmed through the point cloud density in the local point cloud data block, the core point in the core block is searched, the clustering processing time is shortened, the algorithm complexity is reduced, and the operation efficiency of 3D target detection is improved.

The invention relates to the technical field of furniture board blanking, solves the problems of low material utilization rate and high algorithm complexity caused by board texture direction constraint and'one-step cutting 'process limitation in the prior art, and provides a three-stage mixed heuristic blanking method for panel furniture. Comprising the following steps: S1, generating a texture single blank standard length set according to blank data; s2, according to all the single blank block strips, all composite strips with the maximum value under the current condition are obtained through calculation; s3, generating all single blank levels with the maximum value according to the generated single blank composite strip combination; S4, abstracting the single blank levels into a solution combination of a knapsack problem to generate a plate L * W; s5, according to coding of a heuristic sequencing algorithm combined with genetic annealing and decoding calculation of an imprecise positioning algorithm in the three stages, an optimal layout mode under the current condition is obtained; and S6, generating a plate layout drawing according to the optimal layout mode, wherein the plate layout drawing is the optimal plate two-dimensional shearing blanking mode. According to the three-stage mixed heuristic blanking method for panel furniture provided by the invention, through collaborative optimization of the three-stage imprecise layout and the genetic simulated annealing algorithm, an optimal panel layout mode with a high utilization rate can be obtained; and meanwhile, the texture direction constraint in furniture artificial board blanking is ingeniously solved through a blank and strip rotation limiting method, and the operation time of the algorithm can be shortened through a grading solving mode combining a standard size set and three-stage imprecise layout.

The invention belongs to the technical field of signal processing, and discloses an unmanned aerial vehicle sensing task allocation method based on a communication and sensing integrated base station, which comprises the following steps: acquiring a macro base station sensing situation; the macro base station evaluates and calculates the sensing time resource consumption of the sensing integrated base station; the macro base station completes task allocation scheme characterization and problem modeling; the macro base station optimizes an unmanned aerial vehicle perception task allocation scheme; the sensing integrated base station senses the unmanned aerial vehicle and obtains the low-altitude situation of the unmanned aerial vehicle. According to the method, the communication perception common front end and time-division working integrated base station is utilized, the unmanned aerial vehicle perception task allocation method with multi-base-station cooperation is optimally designed, the low-altitude unmanned aerial vehicle perception problem is effectively solved, the time resource utilization rate is greatly improved, and meanwhile, compared with an offline optimal scheme with long time consumption, the method has the advantage that the method is easy to implement. Under the condition that performance is hardly lost, algorithm complexity is greatly reduced, and the requirement for real-time task allocation is met.

The invention provides a torsional vibration self-adaptive control method in a motor driving process of a planetary multi-gear hybrid power system, which realizes vehicle torsional vibration suppression and gives consideration to multi-working-condition adaptability, anti-interference capability and algorithm complexity requirements, and comprises the following steps of: constructing a torque reconstruction and feedback control algorithm; the initial torque demand of the motor is corrected into a gently changing torque, the torque is split into sub-torques with a specific phase difference, the sub-torques are superposed to obtain a reconstructed torque, the reconstructed torque is combined with a feedback torque, and a self-adaptive adjustment torque is output; setting a multi-working-condition torque demand, performing off-line optimization on control parameters, obtaining a constructed working condition control parameter library, performing identification and weighted calculation on a seating point of an initial motor torque demand, and outputting an optimized control parameter to a torque reconstruction and feedback control algorithm; and performing algorithm interference adaptive compensation, obtaining an interference estimated value according to the rotating speed of the output shaft and the nominal inverse model, calculating an interference compensation value through a filter, adding the interference compensation value with the adaptive adjustment torque to obtain a total target torque of the motor, and outputting the total target torque to a controlled vehicle.

The invention provides a camouflage target detection method based on Mangbar capsule routing. A target detection network based on Mangbar capsule routing is constructed, the target detection network comprises an encoder module, a Mangbar capsule generation module, an EM routing layer, a space capsule recovery module and a decoder module, the Mangbar capsule generation module adopts a brand-new capsule generation mechanism, and the space capsule recovery module adopts an EM routing mechanism. According to the method, global information construction is achieved, meanwhile, implicit retention of the context relation of the capsule sequence is completed, and type-level capsules without spatial resolution can be generated, so that subsequent EM routing is changed from a pixel level to a type level, and the algorithm complexity is greatly reduced. According to the method, the parameter quantity and the calculation complexity of the original capsule routing can be greatly reduced, and the camouflage target detection effect is optimized while the detection speed is improved.

The invention discloses a low-complexity channel estimation method and device based on an MIMO system, and the method comprises the steps: recovering a complete matrix through an approximate low-rank matrix, carrying out the singular value decomposition of the complete matrix, building a large-scale spatial multiplexing MIMO system model, carrying out the conversion processing of a received signal on a kth antenna, and obtaining a transmission signal estimation value, and performing channel matrix recovery on the transmission signal estimation value by using an approximation algorithm to obtain a channel matrix estimation value, finally performing signal detection on the channel matrix estimation value by using a zero-forcing technology, and performing iterative updating to obtain an optimal transmission signal estimation value, thereby effectively reducing the algorithm complexity and effectively reducing the calculation amount. The estimated channel matrix is effectively equivalent to the complete channel matrix, and the channel matrix is updated through iterative updating, so that the dynamic change of the channel can be effectively tracked, and the final demodulation signal precision is ensured.

The invention discloses a chip layout image generation method based on a boundary tracking scanning line algorithm. A sub-graph-polygon data structure is designed, a maximum matrix of adjacent polygons of a sub-graph is established, the boundary scanning line algorithm is improved through the method, a complete chip layout is rapidly cut into a sub-graph, and a corresponding binary image is generated. According to the algorithm, a sub-graph circumscribed frame-polygon circumscribed frame data structure is designed, and polygons which are likely to intersect with the sub-graphs are quickly positioned. The data are stored to construct the maximum bounding box of all polygons contained in the current sub-graph, so that the algorithm does not need to calculate the intersection points of the polygons and the sub-graphs, and the time consumption for calculating the intersection points by the algorithm is reduced; in addition, a boundary tracking mode is designed, a scanning line algorithm is improved, intersection points of scanning lines and polygon boundaries do not need to be calculated when a certain line is scanned by storing all points in the boundaries, and algorithm complexity is further reduced.

The invention relates to the field of radio frequency power control, and discloses a radio frequency power control method and system for a 5G mainboard, and the method comprises the steps: 1, capturing the phase fluctuation of a millimeter wave frequency band through employing a superconductive quantum bit array, building a virtual channel response based on quantum entanglement state measurement, generating a channel topological graph, calculating and predicting a channel state based on the channel topological graph, and carrying out the calculation of a predicted channel state; calculating a prediction error according to the measured actual channel state, and adjusting the radio frequency power; 2, fusing the quantum measurement matrix and the deep residual network, introducing a residual attention mechanism to reconstruct an abnormal channel, carrying out local optimization, solving asymmetric Nash equilibrium, reducing algorithm complexity and calculation complexity, dynamically controlling radio frequency power, suppressing interference and improving power amplifier performance; and step 3, analyzing the influence of environmental factors on system errors, jointly calibrating and correcting the errors, constructing a learning model and expanding a dynamic range, determining the actual maximum transmitting power and distributing radio frequency power.

A method for estimating typhoon structural parameters by fusing spaceborne GNSS reflectometer data and radiometer data, which relates to the technical fields of data fusion and typhoon monitoring. The method comprises: acquiring GNSS-R reflectometer data and radiometer data; filtering typhoon data from the GNSS-R reflectometer data and the radiometer data to obtain filtered GNSS-R reflectometer data and radiometer data; performing data spatio-temporal matching on the filtered GNSS-R reflectometer data and radiometer data to obtain a spatio-temporally matched region; performing data fusion processing on the spatio-temporally matched region to obtain fused data; and performing typhoon structural parameter estimation on the obtained fused data to obtain structural parameters of a typhoon at each development stage. Fused data of spaceborne GNSS-R reflectometer data and radiometer data is used for typhoon structural parameter estimation, such that the temporal resolution and accuracy of parameter estimation can be improved, and the algorithm complexity is low.

The application discloses a radar plot and AIS track adaptive parallel matching method and system, and belongs to the technical field of radar and navigation information. In order to reduce the algorithm complexity and the dependence of AIS data, the application comprises the following steps: identifying the database arrival data, performing voting on the projected radar plot in a parameter space by a system error extraction module based on an error distribution function to obtain a plurality of counters and filtering the counters by a filter, finally accumulating the results of all the filters and performing peak value detection to extract the system error of the radar plot; performing speed correlation checking on the searched radar track point by a radar plot index module, and searching the radar track point with a speed parameter meeting the requirements; performing sum and optimization on an error function according to a PARZEN window based on the radar track point with the speed parameter meeting the requirements by an error function optimization module; and updating the membership ship number in the AIS table of the database by using the radar track point with the speed parameter meeting the requirements.

The invention discloses a feature recognition and path planning method for point cloud three-dimensional-two-dimensional-three-dimensional conversion of a curved-surface workpiece, and the method comprises the steps: obtaining three-dimensional point cloud data of the curved-surface workpiece, and carrying out the preprocessing of the three-dimensional point cloud data; projecting and converting the preprocessed three-dimensional point cloud data into two-dimensional point cloud data by adopting a slice projection method, and forming a fitting curve based on the two-dimensional point cloud data; identifying and obtaining a two-dimensional inflection point set meeting the requirement by a curve inflection point detection method based on an angle integral window; converting the two-dimensional inflection point set back to a three-dimensional coordinate space; adopting a two-stage fitting strategy to identify and obtain an optimal fitting plane from the three-dimensional inflection point set; and a plane cutting method is adopted to obtain a planned path for actual machining. According to the method, a complicated point cloud feature extraction problem is converted into a two-dimensional analysis problem with higher calculation efficiency, so that the algorithm complexity is reduced; the problem of low feature extraction accuracy in a noise environment is effectively solved; the robustness under the condition that the outliers exist is enhanced; an intermediate link of CAD modeling in a traditional method is avoided, and accumulative errors are reduced.

The invention belongs to the technical field of electronic countermeasure and target positioning, and relates to a TDOA radiation source tracking and trajectory prediction method based on EKF. The method comprises the following steps: S1, setting N sensor nodes to measure a TDOA observation value; s2, establishing a radiation source motion state space model; s3, calculating a Jacobian matrix to linearize the nonlinear observation equation; s4, executing an EKF prediction and updating step, and updating state estimation by using a TDOA observation value; s5, predicting a future trajectory based on the current state; and S6, performing loop execution to realize continuous tracking. According to the method, the nonlinear problem of TDOA positioning is solved through the EKF algorithm, high-precision tracking and trajectory prediction of the high-speed moving radiation source are achieved, good robustness and real-time performance are achieved, the algorithm complexity is moderate, and the method is suitable for accurate positioning of the moving radiation source in an electronic countermeasure system.

The present invention belongs to the technical field of matrix converters, and discloses a three-phase four-leg three-level direct matrix converter and an SVPWM modulation method. The topology consists of 24 bidirectional switches and 8 flying capacitors. Each of the four legs A, B, C, and N includes 9 output switching states and 6 output voltages. There are a total of 6561 switching states, and the output phase line voltages altogether include six output voltages. The present invention proposes a novel SVPWM modulation method. This method mainly transforms the control of the output voltage from the conventional αβ coordinate system to the abc coordinate system, while the input current is still modulated separately in the αβ coordinate system. The biggest advantage of using the abc coordinate system is that through coordinate translation, the three-level modulation can be simplified to two-level modulation, greatly reducing the increase in algorithm complexity caused by the increase in the number of levels.

The invention relates to the technical field of resource allocation, in particular to an object recognition and task processing method and system for multi-machine cooperation of competition robots, and the method comprises the following steps: pre-defining a plurality of service quality levels of a target object recognition and path planning calculation task of a cooperative robot, and obtaining CPU resource consumption and task processing performance corresponding to each grade based on the visual sensor data resolution and algorithm complexity under different grades. According to the method, a plurality of task service quality levels are defined in advance, and different visual data resolutions and algorithm complexity are mapped into CPU resource requirements and task processing performance indexes, so that quantitative association of tasks and resource requirements is realized, and the resource allocation accuracy is improved; and meanwhile, the CPU load state of the robot cluster is sensed in real time, a dynamic QoS regulation and control instruction is established by combining the load change trend and the task emergency degree, and the real-time performance and sensitivity of resource allocation are improved.

The invention provides a communication and computing resource joint allocation algorithm based on a diffusion model, and relates to the technical field of allocation algorithms. The communication and computing resource joint allocation algorithm based on the diffusion model comprises the following steps: S1, defining input and output; s2, constructing an environment information data set; s3, preprocessing the data set; s4, training an environment information classifier; s5, training a diffusion model; s6, generating a communication and computing resource joint allocation scheme by using a diffusion model; and S7, evaluating a resource allocation scheme. According to the method, real-time Internet of Vehicles environment information is input into a pre-trained environment information classifier to predict the category of the environment information, and the category of the environment information is used as a guide word to guide a diffusion model to generate a proper communication and computing resource joint allocation scheme in the environment. The problems that an existing algorithm is high in complexity, low in convergence speed and prone to falling into a local optimal solution are solved.