356 results about "Weight distribution" patented technology

The invention relates to the technical field of unmanned aerial vehicle flight path planning, in particular to an unmanned aerial vehicle real-time path planning system and method based on dynamic weight distribution and multi-source data fusion. The system comprises a multi-source data fusion module, an integrated laser radar, a millimeter wave radar, a visual sensor and a Beidou positioning unit. The dynamic weight distribution module dynamically adjusts the weight coefficient of each sensor according to the environmental complexity, the threat level and the state of the unmanned aerial vehicle by adopting a mixed decision-making mechanism combining fuzzy logic and reinforcement learning; an improved RRT * algorithm and a Markov decision process are built in the real-time path planning module, and a global optimal path and a local obstacle avoidance track are generated by adopting a layered planning architecture; the unmanned aerial vehicle cooperative control module comprises a dual-redundancy flight control system and a dynamic obstacle avoidance unit; and the communication relay module supports 5G and low-orbit satellite dual-mode communication, updates an environment cognitive model of each unmanned aerial vehicle through federated learning, and realizes multi-source fusion real-time path planning based on dynamic weight distribution and the unmanned aerial vehicles.

The invention provides an inertial navigation attitude resolving method and device based on sensor fusion. The method comprises the following steps: carrying out multi-clock domain synchronous calibration processing on a double-antenna GNSS system, an IMU inertial measurement unit and an antenna servo system in a vehicle-mounted communication-in-motion system; according to the time synchronization reference, carrying out quality evaluation and weight distribution processing on the GNSS satellite signal under the satellite communication in motion antenna pointing constraint; and carrying out multi-sensor fusion resolving processing on the inertial attitude of the vehicle body according to the time synchronization reference and the GNSS observation data weight distribution result. According to the method, millisecond-level time synchronization is realized by establishing a robust extended Kalman filtering model of a multi-dimensional extended state vector, GNSS signal quality is optimized by adopting an adaptive weight distribution strategy of antenna pointing constraint, and attitude fusion precision is enhanced by utilizing high-precision angle feedback of an antenna servo system. The problem that a traditional attitude resolving method in a vehicle-mounted communication-in-motion system is not high in precision is solved.

The invention discloses a tunnel ventilation control system and method based on artificial intelligence, and relates to the technical field of ventilation control, and the method comprises the steps: in a tunnel design stage, building an unsteady-state computational fluid dynamics model based on tunnel three-dimensional linear parameters, calculating turbulence structures under different traffic conditions through a large eddy simulation method, and calculating an unsteady-state computational fluid dynamics model; determining an optimal space configuration scheme of the fan group according to the distribution of the velocity field and the pressure field; in the tunnel construction stage, multiple types of sensor arrays are arranged along the vault and the side wall of a tunnel in a layered mode; in the tunnel operation stage, real-time vehicle tracks and speed distribution information of a traffic monitoring system are obtained; establishing a ventilation demand dynamic prediction model based on space-time correlation analysis, constructing a fan cooperative control model considering airflow organization optimization, and solving an optimal operation strategy by adopting a multi-target adaptive weight distribution algorithm; when a fire characteristic signal is monitored, the multiple sets of fans are coordinated to form a relay type smoke exhaust airflow organization.

The invention relates to the technical field of control systems, in particular to an unmanned aerial vehicle flight control multi-sensor attitude confidence calculation method, which comprises the steps of multi-source sensor data acquisition, sensor independent attitude calculation, confidence coefficient dynamic evaluation, multi-modal data fusion, attitude error compensation and optimization and real-time verification and fault-tolerant processing. Compared with a fixed weight distribution scheme in the prior art, the method has the advantages that the attitude calculation precision is limited possibly, and particularly when the sensor data quality is influenced by environmental factors, the scheme adopts a strategy of checking the data quality in real time, judging the environmental state and dynamically distributing the weight; the validity of sensor data is effectively identified by verifying whether the acceleration modulus length is close to the gravitational acceleration or not and whether the magnetic field intensity is stable or not; meanwhile, the weight of each sensor is flexibly adjusted according to different environment states such as static, maneuvering or magnetic interference of the unmanned aerial vehicle, so that the self-adaptive capability and accuracy of attitude calculation are remarkably improved.

The invention provides a multi-dimensional driver capability evaluation and intelligent matching scheduling system, and the system comprises a data collection module which is used for obtaining driver driving behavior data, vehicle state data and environment data in real time; the preprocessing module is used for carrying out noise filtering, missing value filling and standardization on the acquired data; the multi-dimensional capability evaluation module is used for calculating a driving safety score, an efficiency score and an emergency response score of the driver through a dynamic weight distribution algorithm based on the preprocessed data; the demand analysis module is used for analyzing the route complexity, the time sensitivity and the special service demand of the passenger order; the matching scheduling module is used for generating a matching result according to the driver ability score and the passenger demand and outputting a scheduling instruction; the dynamic optimization module monitors the driver state and the road condition change in real time and adjusts the matching weight; and the interaction module is used for pushing real-time scheduling information and abnormal event early warning to the driver and the passenger. The scheduling efficiency and safety can be improved, and the passenger travel experience and the operation management level are improved.

The invention belongs to the field of battery monitoring, and particularly relates to a battery health state on-line evaluation system considering a temperature-current coupling effect, which comprises an acquisition module, a coupling module, an on-line evaluation module and an early warning module, the acquisition frequency and the window length of current and temperature data are adaptively adjusted through reinforcement learning and a federated frame in the acquisition module; the coupling module constructs a consistency interval based on the current and temperature slope sequence, calculates a segmented current thermal stress factor and a segmented temperature gradient factor, and further generates a temperature-current coupling feature vector; the online evaluation module performs health state evaluation and life decay rate calculation in combination with dynamic weight distribution and principal component analysis; the early warning module gives an alarm in a grading manner according to an evaluation result, and feeds back an attenuation rate to the front-end acquisition module and the online evaluation module to realize closed-loop optimization; according to the method, the accuracy and reliability of battery health state evaluation of the electric two-wheeled vehicle and the tricycle under complex working conditions are effectively improved.

The invention discloses a cooperative steering control method considering man-machine mutual reliability, and aims to solve the problem of man-machine conflict in an existing man-machine co-driving system. The invention relates to the field of intelligent driving. The system specifically comprises an expected path module, a man-machine steering conflict evaluation module, a driver experience evaluation module, a weight distribution module, a non-cooperative game module and a man-machine co-driving vehicle. The expected path module generates an expected path between the driver and the controller; the man-machine steering conflict evaluation module outputs the trust degree of the driver to the controller based on the steering action of the driver, the vehicle state and the expected path of the driver and the controller; the driver experience evaluation module outputs the driver trust degree of the controller based on the vehicle state; the weight distribution module is used for outputting a driver control weight; and the non-cooperative game module outputs an optimal turning angle in combination with the expected path of the driver and the controller and the control weight of the driver, and controls the man-machine co-driving vehicle to complete the cooperative steering action.

The invention discloses a ship power system optimization control method and system based on a simulated annealing algorithm, and relates to the technical field of ship power control, and the method comprises the steps: obtaining operation parameters and historical energy consumption data in real time, and constructing a multi-objective optimization function of dynamic weight distribution; generating an initial temperature parameter and a solution set in combination with the navigation state and the environment data; a neighborhood search strategy disturbance solution set is improved, a new solution is evaluated by using a dynamic acceptance probability function, temperature parameters are adaptively adjusted for iterative optimization, and an optimal control parameter combination is output; and an adjustment instruction set is generated after multi-dimensional efficiency verification, and a propulsion device, a generator set and an energy storage module are cooperatively controlled, so that global energy consumption optimization is realized. According to the method, through data driving and intelligent algorithm fusion, energy efficiency and environmental adaptability are improved, and stable operation under complex working conditions is guaranteed. According to the ship power system optimization control method and system based on the simulated annealing algorithm, the energy efficiency level and the environmental adaptability of the ship power system are improved.

The invention discloses a method and a system for evaluating the reliability of a contact network system under an icing galloping condition. The method comprises the following steps: collecting icing galloping data of the overhead line system in real time, and carrying out load conversion and statistics on parts of the overhead line system in icing galloping by adopting a rain flow counting method; the reliability of parts under the icing galloping condition is calculated through the Miner cumulative damage theory, and weight distribution of the overhead line system is conducted through an analytic hierarchy process; based on the reliability and the weight distribution result of each part under the icing galloping condition, evaluating the reliability of the overhead line system of the data acquisition anchor section until the evaluation task is finished; the method and the system provided by the invention are suitable for evaluating the operation reliability of the contact network system in the icing galloping section, can provide support for operation and maintenance of the contact network system after icing galloping occurs, and have wide application prospects.

The invention relates to a helicopter rotor crack fault identification method and device based on video semantic segmentation. The method comprises the steps that video collection and denoising and stability enhancement processing are carried out through an unmanned aerial vehicle; through designing a global-local feature interaction double-branch network, parallel computing and bidirectional fusion are carried out on context global feature extraction branches and detail mining branches, so that rotor wing crack feature extraction is realized; designing a hierarchical crack feature reconstruction decoder to realize cross-scale fusion and spatial precise alignment of crack features; and designing a multi-frequency-domain boundary sensing enhanced training head, and realizing multi-scale feature extraction and dynamic weight distribution of the crack edge through the synergistic effect of a multi-stage wavelet frequency domain decomposition module and a self-adaptive boundary weighting supervision module. Through the innovative modules, high-precision and robust crack detection is realized aiming at the problems of view limitation, global-local feature fusion difficulty, detail loss caused by down-sampling, edge blur, class imbalance and the like in helicopter rotor crack detection.

The invention provides a low-altitude flight accident intelligent identification method based on multi-source data fusion, and the method comprises the steps: carrying out the encryption processing of flight control data, dynamically mapping environment data through an airspace grid mapping model, carrying out the time-space correlation of the two kinds of data, and obtaining multi-source data; constructing a five-dimensional evidence evaluation system through the information entropy of the multi-source data, and carrying out dynamic weight adjustment on the evaluation system according to a real-time data triggering condition to obtain a causal contribution degree; an accident liability contribution degree is obtained through a weight modeling algorithm; then evidence conflict detection is carried out to obtain recombination credibility; and finally, training a dynamic weight distribution model by recombining credibility, and identifying a low-altitude flight accident to obtain a responsibility division result. According to the method, multi-dimensional data are integrated in real time, a complete flight accident evidence chain is constructed, and a dynamic evaluation and multi-party collaborative management mechanism is introduced, so that a scientific basis is provided for responsibility division, and the responsibility affirmation efficiency and credibility are improved.

The invention discloses a multi-source heterogeneous data flow real-time fusion and weight adaptive correction system and method, and relates to the technical field of multi-source data fusion. Comprising a path monitoring data extraction module, an initial fusion weight determination module, an abnormal transportation event analysis module, an initial correction strategy table generation module, a response updating module and a real-time weight adjustment fusion module. According to the method, a multi-dimensional data quality scoring system is established, so that strong association between weight distribution and data quality is ensured, and fusion precision loss caused by'fixed weight 'is avoided. For an abnormal monitoring point, through indexes such as a data missing rate, a deviation change rate, an auxiliary positioning deviation rate and the like, an abnormal data source type is accurately identified, a weight adjustment scheme of'target data source decreasing amplitude + increasing amplitude data source equal proportion complementation 'is formulated until abnormity is eliminated, and closed-loop correction of'abnormity positioning-weight adjustment-error elimination' is realized.

The invention provides a cross-coupling cooperative control method for double permanent magnet synchronous motors of an aviation airborne suspension lifting system. The cross-coupling cooperative control method comprises the following steps: acquiring a global instruction and real-time state feedback of the double motors; based on the global instruction and the real-time state feedback, constructing a dynamic topology and dynamically adjusting the weight to obtain the weight of the dynamic topology; generating a collaborative correction based on the weight of the dynamic topology and a consistency protocol; dynamic error balanced distribution under load disturbance is carried out based on the collaborative correction amount; and controlling the double motors based on an error balanced distribution result. According to the method, dynamic weight distribution and dynamic error balanced distribution are combined, multi-source interference working conditions such as mechanical unbalance loading, dynamic coupling and parameter disturbance can be dealt with, synchronous errors and disturbance fluctuation in a dual-motor system can be accurately monitored and coordinated, and high-precision stability of cooperative driving is guaranteed.

The invention discloses an intelligent road traffic risk early warning method and system based on vehicle-road cooperation, and relates to the technical field of traffic early warning, and the method comprises the steps: unifying real-time collected road data to a UTM coordinate system through combining time alignment and coordinate conversion, and constructing standardized spatio-temporal data; based on the standardized spatio-temporal data, using a hypergraph neural network to construct a dynamic hypergraph, obtaining node risk features, and generating a comprehensive risk index; according to the risk index feature weight distribution vector, risk root cause probability distribution is obtained through intervention calculation, and a risk level and an early warning type are output; real-time road data are safely aggregated through federal learning, and the dynamic hypergraph is updated in combination with differential privacy. According to the method, the dynamic hypergraph is constructed, the hypergraph neural network is used for capturing the high-order interaction relation between the vehicles, the problem that complex space-time interaction modes between the vehicles are difficult to capture is solved, and the extraction efficiency of node risk features is improved.

The invention discloses a ship dynamic collision avoidance decision-making method based on an improved DWA algorithm. The ship dynamic collision avoidance decision-making method comprises the steps of constructing sea chart environment data containing obstacle density through a grid method; establishing a dynamic elliptical ship domain based on a quaternion ship domain (QSD) model, and calculating a distance collision risk degree (NCRI) between ships based on the dynamic elliptical ship domain; improving a DWA algorithm evaluation function, and adding a path yaw evaluation function and a collision avoidance rule evaluation function on the basis of optimizing a speed and azimuth angle evaluation function; fusing the multiple evaluation functions by adopting a weight distribution method; and finally, generating a local collision avoidance optimization route based on environment data, NCRI and an improved DWA algorithm. According to the method, safety boundary dynamic adjustment, rule compliance constraint and ship control characteristic cooperation are realized, and the robustness and practicability of collision avoidance decision making under complex sea conditions are remarkably improved.

The invention relates to the technical field of vehicles, in particular to an intelligent cabin man-machine cooperative control method and device, a vehicle, a medium and a product. Performing fusion processing on the voice signal data, the visual signal data and the tactile signal data of the current driver by using a preset dynamic weight distribution model to obtain a multi-modal fusion control instruction; and under the condition that the multi-mode fusion control instruction meets a preset sight-voice dual-mode condition, synchronously issuing the multi-mode fusion control instruction to a plurality of target execution devices in the intelligent cabin so as to drive each target execution device to perform cooperative control. Therefore, by constructing a cooperative control mechanism of multi-mode sensing and dynamic weight distribution, the problems that the false wake-up rate is high, the voice and graphical interface cooperative efficiency is low and the interaction mode cannot adapt to a complex driving environment due to a single or separated interaction mode in the related technology are effectively solved.

The invention relates to the technical field of artificial intelligence and remote sensing information processing, in particular to an unmanned aerial vehicle-satellite remote sensing data fusion method based on dynamic weight distribution, which comprises the following steps: step 1, dynamic feature extraction and environment perception; 2, dynamic weight distribution and fusion execution: forming a feature vector by the feature value and an environment interference coefficient, inputting the feature vector into a pre-training weight distribution model, and outputting real-time weights of the unmanned aerial vehicle and the satellite; generating an interpolation scaling factor according to the resolution ratio, and performing up-sampling on the low-resolution data by adopting edge sensing interpolation; pixel-level weighted fusion is carried out on the unmanned aerial vehicle data and the up-sampling satellite data according to the real-time weight; and step 3, performing closed-loop feedback optimization. According to the method, the accuracy, timeliness and quality of remote sensing data fusion are effectively improved in a dynamic weight distribution and environment perception mode, and the method has wide application value.

The invention belongs to the field of road vehicle control systems, and relates to an automatic driving takeover method and system based on cognition-execution double-layer decision. The system comprises a multi-mode large model, a takeover mode judgment module, an early warning intensity calculation module, a sensing reminding module, a preparation state evaluation module, a self-adaptive weight distribution module, a control signal output module and a guide interface generation module. Wherein the multi-modal large model is used for reasoning according to input data and outputting structured takeover decision information; the early warning intensity calculation module is used for calculating takeover early warning intensity changing along with time; the preparation state evaluation module is used for evaluating the driver's preparation for taking over; the control signal output module is used for realizing smooth transfer of the man-machine control right under the safety constraint through a control fusion mechanism according to the self-adaptive control weight of the driver; the system not only improves the accuracy and interpretability of the takeover decision, but also effectively improves the takeover performance and reduces the takeover risk.

The invention belongs to the technical field of airport clearance safety, and discloses an airport clearance ultrahigh ground feature extraction method fusing LiDAR point cloud and a panoramic segmentation model, and the method comprises the steps: collecting point cloud data in an airport clearance protection region through employing an unmanned plane airborne LiDAR, and carrying out the noise filtering, outlier elimination and datum point-based elevation standardization; constructing a data set with semantic tags and instance numbers, inputting the data set into a SuperCluster panoramic segmentation model, sequentially executing feature extraction, adjacency relation construction, edge weight distribution, local supervision and super-point aggregation, and outputting a segmentation result with semantics and instances; extracting a highest point in each instance point set, and performing spatial comparison with an inner horizontal plane, an approach plane and a transition plane limiting plane established according to an ICAO standard to obtain an instance category, a highest point coordinate, an intrusion depth and a spatial relationship; and performing three-dimensional visual display on the judgment result through the clearance supervision platform.

The invention discloses a vehicle control method and device based on multi-source information fusion and actuator health state monitoring. The vehicle control method comprises the steps that driver state data, vehicle dynamics data and environment information data are obtained; sensing and fusing the multi-source data to generate comprehensive sensing data; performing scene classification according to the comprehensive perception data, identifying and judging a current driving scene type, and determining a control demand of current driving; health state monitoring is conducted on vehicle actuators, and the working performance and reliability of all the actuators are evaluated in real time; according to a scene judgment result and an actuator health state monitoring result, dynamically adjusting weight distribution of each control target in a decision making process; and based on the adjusted weight distribution, executing multi-target collaborative optimization, generating an optimal control strategy, converting the optimal control strategy into a control instruction, and issuing the control instruction to a vehicle execution mechanism. And the technical bottlenecks of the existing system in complex scene adaptability, actuator collaborative optimization, fault handling capacity and the like can be solved.

The invention discloses a mining area roadway stability intelligent classification and support scheme design method, and the classification method comprises the steps: carrying out the denoising through employing a denoising method based on statistical deviation, constructing a weight distribution frame based on a random forest algorithm, carrying out the feature importance evaluation of each feature variable through employing a decision tree integration method, and carrying out the design of a support scheme. Quantifying the influence proportion of each characteristic variable on the stability of the mining roadway, and generating a weight matrix containing the influence proportion of each characteristic; and constructing a mining area roadway stability intelligent classification model based on a support vector machine, adopting a radial basis function as a kernel function to map a feature variable to a high-dimensional space, adjusting a model classification boundary by iteratively optimizing penalty parameters, and training the model by using historical data to obtain an optimized mining area roadway stability intelligent classification model. Accurate evaluation and intelligent classification of the stability of the roadway in the roadway mining area are achieved, the supporting effect and the roadway safety are effectively improved, and technical support is provided for safe and efficient mining of the mine roadway.

The invention relates to the technical field of vibration tests, in particular to a ground resonance test suspension device and a test method. The ground resonance test suspension device is provided with a bearing frame body, a hoisting mechanism, a bearing mechanism and a suspension mechanism, the bearing frame body is composed of a bearing supporting column, a cross beam and a longitudinal beam, and the position of the cross beam can be adjusted; a bracket of the bearing mechanism is provided with a bottom beam, an auxiliary supporting beam and a supporting beam, and a flexible air cushion is arranged on the supporting beam. A flexible supporting unit of the suspension mechanism is provided with a main spring rope, an upper hanging scaffold, a lower hanging scaffold and an auxiliary spring rope, and an auxiliary spring rope connecting unit can control the state of the flexible supporting unit. The invention further provides a ground resonance test method which comprises the steps of suspension pretreatment, aircraft suspension and test, and the equivalent stiffness of the suspension mechanism can be adjusted. The aircraft suspension system can adapt to different aircraft sizes and weight distribution, and the equivalent stiffness of the whole suspension system can be adjusted to meet the test support frequency requirement.

The invention discloses a driving capability improvement control method and system for an unmanned vehicle. The method comprises the step of realizing road condition perception and tire pressure optimization through vehicle cloud cooperation. A front vehicle collects road surface data and uploads the road surface data to a cloud server after preprocessing, and the server carries out aggregation analysis and constructs and updates a road surface state map. The rear vehicle requests road condition data, verification is carried out, remote data and local data are fused after weight distribution, and road surface state estimation is generated. And calculating a road surface friction coefficient and a weather influence factor based on the fused data, processing tire temperature data, monitoring a load-carrying vehicle speed, formulating an adjusting strategy, and carrying out modeling solving to obtain the optimal tire pressure. And an adjusting strategy is generated according to the optimal tire pressure, differential tire pressure control is implemented, adjustment is executed through closed-loop and predictive control, the effect is monitored in real time, and data are uploaded to a cloud optimization system. The control stability and trafficability of the vehicle under complex road conditions are effectively improved, and driving safety and riding comfort are guaranteed.

The invention discloses a three-order precision nonlinear interpolation smooth factor design method based on a five-point template, and belongs to the field of computational fluid mechanics, a high-precision format realizes unification of high precision, no oscillation and multi-scale adaptability in complex flow simulation through nonlinear weighting, multi-order combination and smooth factor design, and the method has the advantages of high precision and high precision. The device is an indispensable numerical tool for aerospace. Wherein the smooth factor is a core mechanism for realizing shock wave capture and high resolution in a high-precision format, and weight distribution is dynamically adjusted through smoothness of a quantitative solution, so that balance among shock wave capture, high precision and low dissipation is achieved. Therefore, the invention provides the Z-type third-order precision nonlinear interpolation smooth factor which is based on the five-point template and meets the condition that the optimal precision of the nonlinear weighting framework of the multi-order candidate template is sufficient. A new consistent high-order smooth factor meeting the optimal interpolation precision is designed by using a consistent high-order smooth factor method.

The invention belongs to the technical field of air transportation management, and discloses a multi-source data fused sorting method and system for airport flight coordinated launch. The method comprises the following steps: collecting a multi-source data source of a flight; performing multi-source data fusion, generating a five-dimensional dynamic situation matrix, and reflecting the comprehensive state of the flight in real time; judging whether manual priority ranking processing is carried out or not; by constructing an intelligent decision-making system for aircraft departure sorting, optimization target decision-making and intelligent sorting are carried out; the visual sorting interface displays the alternative push-out sorting schemes; obtaining a sorting result of flight coordination deduction; establishing a dynamic reentry mechanism for flights with updated resources or changed states, and reinserting the flights into the global sorting queue; and balancing the adaptation degree of the original priority and the new state of the flight through a dynamic weight distribution algorithm. The airport flight scheduling efficiency is improved, so that flight delay is reduced, the operation efficiency is improved, and the passenger satisfaction degree is enhanced.

The invention discloses a stacking loading and unloading structure of a stacking machine, comprising: a main body unit comprising a machine body on which two groups of telescopic forks are arranged; the adjusting unit comprises a fixed seat fixedly installed at the telescopic end of the telescopic pallet fork, two sliding parts and a detection unit fixedly connected with the sliding parts are symmetrically arranged on the fixed seat, and the detection unit is controlled to measure the weight of goods at the two ends of the fixed seat through displacement of the two sliding parts; and the limiting assembly comprises a second air bag arranged on the fixing seat and a plurality of clamping pieces inserted into the fixing seat in a sliding mode, and the limiting assembly is used for reinforcing connection between the fixing seat and the freight plate. Through the detection unit and cooperation of the two sets of sliding parts and the detection unit, in the process that the telescopic pallet fork lifts up the freight plate, the weight distribution difference of goods on the freight plate is detected in real time, the position of the telescopic pallet fork relative to the freight plate is judged and adjusted in time, and the situation that due to the unstable gravity center, the goods slide, incline or fall off in the transportation process is avoided.

The invention relates to the technical field of intelligent loading, and discloses a particle swarm optimization-based three-dimensional loading optimization method and device and a medium, and the method comprises the steps: defining the size and volume of a carriage, generating tray data, and extracting the arrangement sequence and grouping information; calculating the plot ratio of the trays after loading, and analyzing the position vector to obtain a tray arrangement sequence and a grouping scheme; obtaining a single group fitness and an overall fitness; and constructing a particle swarm, initializing particle positions and speeds, then defining position vectors, updating an individual optimal solution and a global optimal solution, and outputting an optimal loading scheme. According to the method, multi-dimensional constraints such as spatial layout, weight distribution and loading and unloading sequence of cargo loading can be effectively fused, and global optimization of a loading scheme is realized. Through a particle swarm intelligent search mechanism, the arrangement and grouping strategy of trays is continuously iteratively optimized, the loading efficiency and the space utilization rate are remarkably improved, meanwhile, the stability and safety in the cargo transportation process are guaranteed, and the good practical application value is achieved.

The invention discloses a traffic accident video evidence analysis system and method based on multi-modal deep learning, and relates to the technical field of deep learning, and the method comprises the steps: carrying out the target recognition and tracking through radar point cloud, extracting a target number component and a motion variance component, carrying out the calculation through normalization and linear combination, and obtaining a dynamic fusion weight of each modal; and quantization and adaptive adjustment of scene complexity are realized. Then extracting each modal feature vector in a preset time sequence window, and performing weighted fusion based on the weight to form a unified feature vector; and inputting the uniform feature vector into a pre-trained accident classification model, and outputting whether an accident occurs or not and a type result. The method has the advantages that through complementary enhancement of multi-modal data, the accident identification accuracy in complex scenes such as low illumination and rain and fog is improved; adaptive adaptation to different scenes is realized through dynamic weight distribution; the whole scheme forms a closed loop, and has high robustness and practical value.

The invention discloses a deep reinforcement learning highway variable speed limit control method based on multi-target dynamic weight distribution, and the method comprises the steps: 1, collecting the traffic flow information of a road section, and obtaining the traffic state data; 2, defining an action parameter and a reward function of the deep reinforcement learning method; 3, constructing a network model of deep reinforcement learning, and establishing a playback memory bank storage sample; and 4, reward function weight updating is carried out based on an entropy weight method, an optimized control strategy is obtained based on a sampling training deep reinforcement learning network model, and speed limit control is carried out on each control unit. According to the method, dynamic weight distribution of the traffic efficiency and the safety reward function is achieved, the method can adapt to a complex traffic environment, meanwhile, the operation efficiency of the expressway is effectively improved, and the safety risk is reduced.

The invention discloses a power distribution overhead line icing measurement method, and belongs to the technical field of icing measurement, and the method comprises the steps: calculating the icing risk probability of a key position coordinate through an icing risk probability model, and constructing a high-precision sensor deployment multi-objective function through the combination of the icing risk probability, and carrying out the high-precision sensor deployment; constructing an asymmetric icing physical model to perform icing state analysis on the fused high-precision sensing data to obtain icing weight distribution and icing direction of the power distribution overhead line; and carrying out real-time early warning feedback on the icing weight distribution and the icing direction of the power distribution overhead line. According to the method, the ice layer weight distribution and the asymmetric direction of each key point of the power distribution overhead line are dynamically calculated by adopting a mode of combining a nonlinear fusion algorithm and a physical and mechanical model, and meanwhile, a risk probability model and a heuristic deployment optimization strategy are introduced, so that high-precision sensor precise layout and dynamic early warning feedback are realized; and the spatial resolution and response efficiency of icing monitoring are remarkably improved.