52results about How to "Resolve distortion" patented technology

The invention relates to the technical field of power system fault detection, in particular to a factory power distribution system fault interval section determining method and system and terminal device.The factory power distribution system fault interval section determining method comprises the steps that when a power failure fault of a power distribution system is monitored, voltage and current data within a preset duration before the fault is obtained; comparing the phase current amplitude with a rated current threshold value, judging that the phase current amplitude is an external fault if the phase current amplitude does not exceed the limit, calculating a zero-sequence current if the phase current amplitude exceeds the limit, and comparing the zero-sequence current with a zero-sequence threshold value to distinguish a ground fault or an inter-phase fault; adaptively calling a phase difference analysis process according to an identification result, calculating a phase difference between a zero-sequence current and a zero-sequence voltage during a grounding fault, calculating a phase difference between a fault phase voltage and a fault phase current during an inter-phase fault, and judging whether a fault interval is internal or external according to a phase difference interval; and a structured message is automatically generated by integrating fault types, interval judgment and triggering criterion paths, so that quick and accurate fault positioning and handling are realized.

The application discloses a double-reference parameter fusion DSQM non-cooperative signal separation method, device, equipment and medium, and relates to the technical field of digital signal processing.The method comprises the following steps: receiving an input signal and performing symbol extension; estimating an original input noise standard deviation, a weak signal and a signal-to-noise ratio, and calculating a right shift number and a dynamic weighted fusion coefficient; performing perturbation injection and phase compensation on the extended signal to generate an actual separation input signal, and estimating a noise standard deviation thereof; calculating an adaptive shift number k of the modified signal-to-noise ratio according to the original input weak signal and the actual separation input signal by using a dynamic weighted fusion algorithm; further calculating a left shift number m; amplifying the actual separation input signal by shifting the actual separation input signal left by m bits; sequentially performing right shift n-bit and left shift n-bit to extract a strong signal component; performing time domain subtraction to separate a weak signal component; respectively shifting m bits to the right to recover the magnitude; then respectively performing symmetric saturation truncation and asymmetric truncation to recover the bit width output. The application improves the signal separation quality.

The present application relates to a kind of low-carbon park integrated energy system multi-agent collaborative carbon responsibility allocation method, steps include: the multi-source data of park is collected, with the minimum integrated energy system total operating cost as goal to carry out multi-energy flow optimization scheduling;With the dynamic correction model of two-way carbon flow considering the carbon flow transfer caused by energy storage charging and discharging process, dynamic carbon flow distribution matrix is constructed;Using load carbon responsibility decoupling allocation model, total load is decoupled into multiple sub-load categories, and different carbon responsibility allocation coefficients are given for each type of sub-load;With each interest party in park as the participant of game, Nash bargaining as collaborative optimization framework, maximum all participant utility gain as objective function, Shapley value as the basis for carbon responsibility allocation, multi-agent collaborative optimization is carried out, and the carbon responsibility allocation scheme of each subject is obtained in combination with user carbon responsibility allocation coefficient.Compared with prior art, the present application ensures the fairness, efficiency and operability of carbon responsibility allocation scheme.

This application presents a method for providing clear, accurate, and intelligent AI representations of complex, multi-element images. It employs an improved Faster R-CNN model as the image processing model, replaces VGG-16 with DenseNet for feature extraction, optimizes bounding box selection through dimensional clustering, and enhances model stability by predicting relative grid cell coordinates. A text generation architecture is constructed based on a hierarchical recursive network. The first layer, a global long short memory layer, determines the topic vector based on the fusion vector to increase the number of descriptive sentences. The second layer, a sentence BRNN layer, generates accurate and rich descriptive sentences. A fully convolutional localization layer effectively integrates and transfers block features, constructing an end-to-end architecture that collaboratively optimizes image processing and text generation. This method addresses the problems of high model complexity, inaccurate bounding box localization, and insufficient association of descriptive sentences in existing technologies, generating denser, more accurate, and natural descriptive language that conforms to language habits. It is suitable for various application scenarios such as visual intelligent question answering and video subtitle generation.

The application belongs to the technical field of point cloud processing, and discloses a multi-modal semantic mapping method and system based on a master-slave architecture. The method comprises the following steps: S1, extracting feature points in point cloud data and calculating the covariance matrix of the neighborhood point cloud of each feature point; S2, solving the optimal pose corresponding to the minimum of the target function; S3, converting the obtained point cloud data into a world coordinate system using the optimal pose, fusing all frame point cloud data in the world coordinate system to form a three-dimensional point cloud map, projecting the three-dimensional point cloud map to a camera coordinate system, and inputting the three-dimensional point cloud map projected to the camera coordinate system into a semantic segmentation neural network to obtain the semantic label corresponding to each point cloud in the three-dimensional point cloud map, thereby achieving semantic mapping. Through the application, the problems of weak dynamic adaptability, insufficient model generalization ability and long-term running precision decay of the traditional three-dimensional semantic mapping method in a dynamic complex scene are solved.

The application provides a dynamic modeling method and system for multi-field skill AI automatic semantic labeling, and relates to information retrieval, medical health and cross-field skill fusion scene. Through nonlinear dynamic modeling and biomedical stability control, the semantic label automatic generation and real-time matching of talent skill and health intervention effectiveness are realized. The application innovatively integrates the sound wave frequency of fetal brain promoting music method, maternal action instruction and blood type nutrition scheme into the "skill-time" frequency domain space, combines the Gevrey smoothing operator and fractional derivative evolution equation, accurately separates the temporary fetal movement fluctuation and low-frequency core characteristics, and breaks through the characterization limitation of traditional models on nonlinear biological coupling effect. By introducing the "intervention effect-time" frequency domain component and local dependency constraint, the short-term effectiveness and long-term value of music intervention on fetal neurodevelopment are quantified, and the model distortion problem caused by extreme data or cross-field parameter resonance in traditional methods is solved.

The application provides an image fusion method and device and a storage medium, and belongs to the technical field of image fusion.The method comprises the following steps: importing a high-resolution hyperspectral image to be fused, a high-resolution multispectral image to be fused, a plurality of original high-resolution hyperspectral images and an original high-resolution multispectral image; performing down-sampling processing on the original high-resolution hyperspectral images respectively to obtain original low-resolution hyperspectral images; and constructing a training model, wherein the training model comprises a feature analysis network, a target vector field construction network and a predicted vector field construction network.The application effectively avoids the spectral distortion problem caused by the difference in spectral response of a sensor in a traditional fusion method, and overcomes the defects of lack of constraints in the intermediate process and insufficient stability of results in a single-step regression method.

The invention relates to the technical field of data processing, in particular to a BIM model-based multi-party cooperative construction management and control method and system, and the method comprises the steps: obtaining the point cloud data of a real-time construction project, and determining the optimal thick side length of a voxel block according to the down-sampling rate of the point cloud data and the unit side length of the spatial distribution of the point cloud data; voxelization is carried out on the point cloud data based on the optimal coarse side length to obtain a plurality of voxel blocks, and the average density of point cloud distribution of the point cloud data in the voxel blocks is determined; determining the optimal side length of the voxel block according to the average density, the down-sampling rate, the optimal coarse side length and the unit side length, and performing down-sampling on the point cloud data based on the optimal side length to obtain a down-sampling project point cloud set of the real-time construction project; and constructing a BIM model of the real-time construction project by using the down-sampling project point cloud set, and managing and controlling the real-time construction project by using the BIM model. Therefore, the construction accuracy of the BIM three-dimensional model is improved.

The application discloses a kind of blood sampling test tube outer wall laser printing information and bar code device and its using method in mechanical technical field, including base and test tube, and test tube is prefabricated with engraving layer, and test tube holder and laser engraver are arranged on base;Test tube holder includes electric screw, and electric screw is slidably connected with mounting platform, and mounting platform is provided with cylinder, and the output shaft of cylinder is fixedly connected with mounting plate, and mounting plate bottom is rotatably connected with rotating rod, and rotating rod is provided with driver for driving rotating rod to rotate, and rotating rod bottom is fixedly connected with mechanical gripper for grabbing test tube;Electric hatch is opened in base, and material delivery bin for storing test tubes of various colors and controller are arranged in base.The technical scheme of the application can perform laser printing on test tube, and improve label marking efficiency.

The application provides a curved panel defect detection method and system. The method comprises the following steps: acquiring multiple perspective images of a curved panel from multiple perspectives; preprocessing each perspective image, performing projection transformation based on the preprocessed perspective image, and reconstructing a panoramic development map of the curved panel; performing defect recognition and detection based on the panoramic development map; and outputting a defect recognition and detection result of the curved panel. The method can accurately locate and determine defects, and reduces the false detection rate and the missed detection rate.

This application provides a method, equipment, medium, and product for testing power line signals, relating to the field of signal processing technology. The method includes: dividing the power frequency cycle into N test time cycles using a voltage zero-crossing detection unit and generating corresponding N wideband test data packets; sequentially sending the wideband test data packets within M consecutive power frequency cycles to obtain M test signals, which are then collected to form N sub-signal sets; calculating the fluctuation variance of each sub-signal set and generating a confidence coefficient; synthesizing N reference sub-signals and dividing them into high-confidence sub-signals and sub-signals to be repaired based on the confidence coefficients; reconstructing and repairing the sub-signals to be repaired using the channel response characteristics of the high-confidence sub-signals to generate repaired sub-signals; and finally concatenating the high-confidence sub-signals and repaired sub-signals in time sequence to obtain the final test signal. This solution solves the technical problem of how to obtain high-precision power line test signals throughout the entire power frequency cycle.

The invention relates to the technical field of reactors, in particular to a corrosion product deposition growth prediction method, a storage medium and electronic equipment. The corrosion product deposition growth prediction method comprises the following steps: establishing a three-dimensional model, and carrying out mesh generation on the three-dimensional model to obtain a mesh model; obtaining the thickness increment of the corrosion product; converting the thickness increment of the corrosion product on the grid unit surface into a node movement amount by using a node movement algorithm, and driving grid nodes to displace in each time step length according to the calculated node movement amount based on a dynamic grid technology; obtaining an updated grid model; the deposition rate and the erosion rate of the corrosion product are solved on the updated grid model, and the thickness increment of the updated corrosion product is obtained. The application of the invention can improve the safety of the reactor, and has important values for protecting the integrity of the fuel rod, preventing the leakage of radioactive substances and ensuring the safe operation of the reactor.

The invention discloses a power transmission and transformation equipment component mechanism and data mapping model construction method, which comprises the following steps of: 1, establishing a model simplification rule, and simplifying a power transmission and transformation equipment component geometric model; 2, an electrostatic field module, a solid heat transfer module, a solid mechanical module and a physical field coupling module are built based on a finite element analysis platform, and accurate simulation of the multi-physical field coupling process is achieved through parameter setting and data interaction between the modules; step 3, establishing a one-to-one correspondence relationship between simulation data and physical mechanisms by taking the core physical mechanism of the power transmission and transformation equipment element as a link, and forming a structured mechanism-data mapping system; 4, performing mapping verification and iterative optimization to form a closed-loop optimization mechanism; according to the method, calculation precision and analysis efficiency are both considered, a multi-physics field coupling mechanism is integrated, and an accurate power transmission and transformation equipment element mechanism and data mapping model is constructed.

The invention provides a traffic state prediction method and device, a terminal and a medium, and relates to the technical field of intelligent traffic, and the method comprises the steps: recognizing a key traffic road section as a key node based on the historical traffic state data of each traffic road section; constructing a double-layer traffic network structure according to an identification result; a first traffic network structure in the structure comprises all traffic road sections and a spatial connection relation thereof, and a second traffic network structure only comprises key traffic road sections serving as key nodes and a spatial connection relation thereof; and extracting network-level spatial features and key node spatial features based on a double-layer traffic network structure, guiding fusion of the network-level spatial features and the key node spatial features based on prior knowledge information related to each traffic road section, and predicting the traffic state based on the fused spatial features. According to the method, the key node characteristics in the traffic network can be combined, the traffic road segment priori knowledge is utilized to guide feature fusion, and the traffic state prediction accuracy and stability are improved.

The present application relates to the field of intelligent prediction, and particularly relates to a weak adhesion hydrogel adhesion performance prediction method and system, comprising the following steps: S1: obtaining hydrogel material parameters and interface state information and preprocessing; S2: identifying state labels through a classification model to form an extended data set with state labels; S3: constructing a segmented physical model system driven by the interface state to obtain state-adaptive physical prediction results; S4: embedding the state-adaptive physical prediction results into a physically-constrained neural network structure to construct a prediction model and obtain model prediction results; S5: using a multi-model collaborative fusion mechanism, comprehensively considering the state-driven physical model output and the model prediction results, and obtaining the final prediction value through a weighted fusion manner; S6: taking the final prediction results and model internal parameters as inputs, performing feature contribution degree analysis, identifying key influencing factors, and outputting parameter optimization suggestions. The present application realizes high-precision and high-generalization-capability prediction.

The invention relates to the field of slope safety monitoring, in particular to a surface mine slope deformation monitoring method based on Beidou positioning, which comprises the following steps: S10, acquiring a slope radar remote sensing image, and constructing a slope terrain three-dimensional model; and the soft depth of each position of the side slope is obtained through processing in combination with the rock stratum gaps of each position of the side slope, and a soft area and a hard area are marked in the three-dimensional terrain model according to the soft depth. According to the invention, soft and hard areas of the slope are divided through radar remote sensing, Beidou positioning devices are arranged in a targeted and differentiated manner, geological characteristics and terrains are adapted, monitoring resource configuration is optimized, environmental influence and maintenance cost are reduced, data verification of adjacent devices is combined, monitoring precision and credibility are improved, and the method is suitable for large-scale popularization and application. The problem that the error of the acquired deformation monitoring data of the side slope is large when Beidou positioning devices are uniformly arranged on the side slope with obvious geologic structure difference is solved.

This application provides a coal production prediction method based on a detailed 3D model of a coal mining face, relating to the field of coal production prediction. The method includes: collecting boundary data of the exposed dip profile behind the strike of the coal mining face; extracting discontinuity boundary lines from the boundary data; performing polynomial fitting and extension on the discontinuity boundary lines to obtain prediction boundary data for the dip profile to be predicted; constructing a detailed 3D model of the coal mining face based on the prediction boundary data of the dip profile to be predicted; meshing the detailed 3D model of the coal mining face, obtaining the ash content value of each mesh using spatial interpolation, and calculating the average ash content of each mesh; obtaining the number of meshes corresponding to the coal seam and rock strata in the preset mining area, and predicting the coal production based on the corresponding number of meshes. This application enables production prediction to be directly correlated with the spatial occurrence morphology of coal and rock, improving the accuracy and reliability of the prediction.

This invention provides a method, system, equipment, and medium for testing the underwater performance of a submarine's cover layer. The testing method includes: applying shock wave excitation to the cover layer of the submarine under test, and acquiring sediment volumetric strain signals and acoustic pressure signals; extracting the strain-time spectrum and acoustic pressure-time spectrum respectively; extracting strain feature vectors and acoustic pressure feature vectors; and analyzing the correlation coupling characteristics between sediment mechanical compression behavior and acoustic scattering changes through preset frequency domain coupling indices and preset time domain coupling indices, thereby achieving performance testing of the submarine's cover layer. This invention, by simultaneously acquiring sediment volumetric strain signals and acoustic pressure signals related to the cover layer under shock wave excitation and analyzing the correlation coupling characteristics, establishes an intrinsic correlation between mechanical and acoustic responses. This reflects the synergistic relationship between the cover layer's resistance to shock deformation and maintenance of acoustic concealment under dynamic loads, resulting in a more comprehensive and accurate evaluation of its overall underwater performance.

The invention discloses a point cloud map determination method and device, equipment and a storage medium, and relates to the technical field of computers. The method comprises the following steps: acquiring point cloud data, inertial measurement data and GNSS data; based on the point cloud data and the inertial measurement data, utilizing a laser mileage calculation method to obtain corrected point cloud data, first global pose data and relative pose data; determining second global pose data under the GNSS data coordinate system; based on the corrected point cloud data, the relative pose data and the second global pose data, constructing an initial factor graph to determine a relative pose constraint in the window and a relative pose constraint between the windows; determining a GNSS relative pose constraint based on the second global pose data and the GNSS data; and obtaining a target factor graph based on the initial factor graph, the GNSS data, the in-window relative pose constraint, the inter-window relative pose constraint and the GNSS relative pose constraint, and carrying out graph optimization processing on the target factor graph to obtain a point cloud map.

The application provides a method for inhibiting distortion of a VIVALDI antenna pattern of an ultra-wideband, and relates to the field of mobile communication antennas. Two horizontal slots are loaded on a metal radiation surface, one is a long slot, which is used for prolonging the path of an edge current, so that the antenna can be miniaturized to a certain extent, the length of the long slot is 75-90 mm, the width of the long slot is 3-5 mm, and the loading specific position is located at h12 above an antenna resistance loading slot, h12 is 7-12 mm; the other is a short slot which is opened on a current radiation slot of the antenna, and the loading specific position is located at h11 above the antenna resistance loading slot, h11 is 12-22 mm. The length of the short slot is 5-15 mm, and the width of the short slot is 3-5 mm. The application solves the distortion of the antenna pattern in a frequency band, and also has a certain improvement on the standing wave of the antenna.

This invention provides a method for implementing car audio effects based on a real performance venue model. The method includes: S1: placing a sound source to be tested in the target performance venue, and setting up at least two sets of performance venue test points with the sound source to be tested as the origin, collecting the performance venue impulse response at each performance venue test point, and forming an original acoustic signal; S2: processing the original acoustic signal to obtain a standardized acoustic signal; S3: controlling the car audio system in the car cabin to play the test sound, setting up at least two sets of cabin test points in the car cabin, and collecting and obtaining the cabin impulse response; S4: extracting the performance venue impulse response and the cabin impulse response, calculating and obtaining the target impulse response; S5: calculating the target audio signal y(t) by performing a convolution operation between the standardized acoustic signal and the target impulse response; S6: adjusting the sound effect parameters of the target audio signal and playing it through the car audio system in the car cabin.

The application provides a dynamic DR image motion correction method and system based on a video stream, and relates to the technical field of image correction.The method comprises the following steps: calculating the spatial correlation coefficient of a first topographic kernel and a second topographic kernel according to the local gray distribution difference of the two, and constructing a time-varying baseline connecting the first topographic kernel and the second topographic kernel based on the spatial correlation coefficient; when the length of the time-varying baseline in the current frame changes relative to the length of the corresponding baseline in the previous frame exceeds a preset threshold, moving the first topographic kernel until the change falls within the preset threshold to obtain an updated first topographic kernel.The application extracts multiple types of topographic kernels, combines spatial correlation and phase matching to complete dynamic updating of the kernel and transmission of the motion vector, relies on the kernel to construct a motion deformation grid to realize pixel-level image transformation, can effectively compensate for the motion offset in the dynamic DR image, suppresses the motion artifacts, and improves the overall clarity of the video stream image.

The application discloses a kind of visual detection methods for motor shaft, including real-time acquisition relevant parameter of motor shaft, construct the CPS data model of motor shaft, real-time association with digital twin model is carried out to physical detection unit, manufacturing process state and data transformation time sequence are digitalized mapping;Based on the real-time state vector that fusion output is based on multi-source data and digital twin model, carry out adaptive imaging parameter regulation and control and multi-station collaborative imaging planning based on deep reinforcement learning;Based on real-time state vector and imaging parameter, image pre-processing is carried out, and using distortion correction algorithm based on physical model compensates dynamic deformation;Combine multi-station image, using multi-scale feature fusion algorithm detects the tiny defect of motor shaft;Hierarchical attention classification network is constructed, and hierarchical classification is carried out, and the graph neural network of multimodal fusion is simultaneously established to diagnose output for the cause of defect, the application improves the adaptability and precision of motor shaft fast detection imaging.

The utility model relates to the technical field of pile foundation construction, and provides a pile foundation reinforcement cage in-hole fixing device which comprises an air compressor, an air conveying pipe, a valve set, a pressure gauge, an air bag and a guide cage fixing system. The problem that the reinforcement cage cannot be fixed under the special working conditions such as complex geology, limited space and full pile casing follow-up is solved through buoyancy generated by the air bag in mud in combination with rigid constraint of the guide cage fixing system. The guide cage fixing system is formed by welding lifting points, main reinforcements, restraining reinforcements and stirrups, the restraining reinforcements are arranged in an equilateral triangle shape to limit movement of the air bag, and it is guaranteed that the reinforcement cage is erected and the center position is accurate.

The invention provides a cyclic peptide three-dimensional structure prediction method based on a large language model, and belongs to the technical field of crossing of artificial intelligence and computational biology. According to the method, a prediction process is divided into a sequence semantic analysis stage and a space structure generation stage: in the sequence semantic analysis stage, deep semantic understanding is carried out on a cyclic peptide amino acid sequence and structure labeling information by utilizing a pre-trained large language model, and a cyclization site is accurately identified; fusing implicit structure laws such as interaction of amino acid residues and cyclic conformation tendency, and outputting a structured feature vector; and carrying out conformation optimization and atomic-scale detail modeling on the initial three-dimensional skeleton in combination with physical constraints and molecular force field parameters. The method effectively overcomes the problems that a traditional method depends on limited physical and chemical characteristics and is weak in generalization ability on small sample cyclopeptides, significantly improves the accuracy and applicability of cyclopeptide three-dimensional structure prediction, and is especially suitable for high-precision modeling requirements of natural cyclopeptides and artificially designed cyclopeptides in drug research and development.

The application discloses a flattening method and device for a large-curvature composite T-shaped stringer, which comprises the following steps: obtaining a three-dimensional model of the T-shaped stringer, and extracting a rib surface of left and right L-shaped components and an R-angle upper edge line; constructing a spanwise development reference line based on the R-angle upper edge line, discretizing the rib surface into a plurality of cross-section positions along the reference line, replacing a curved surface profile with a straight line segment to construct a single curved approximation surface, and obtaining a flat plate rib profile through a geometric flattening algorithm; based on the profile and a web reinforcement layer information, mapping an R-angle layer-losing boundary to a plane through arc length equivalence to obtain a layer-losing position; extracting a 0-degree reference line corresponding curved surface, discretizing, straight line replacing and constructing another single curved approximation surface along the same reference line to obtain a 0-degree reference line plane profile through flattening; and finally aligning the profile with the flat plate rib profile to obtain a complete flat plate material layer flattening result. The application can improve the flattening precision of the large-curvature composite T-shaped stringer and realize integrated and collaborative flattening.

The application discloses a flight control actuator performance testing device capable of simulating a real service environment, comprising: a test bench, wherein a test channel one and a test channel two are arranged on the test bench; an actuator loading testing mechanism one, which is arranged in the test channel one and is used for realizing force position simulation, variable stiffness simulation and inertia simulation of the actuator; an actuator loading testing mechanism two, which is arranged in the test channel one and is used for realizing force position simulation, stiffness simulation and inertia simulation of the actuator; the actuator loading testing mechanism one and the actuator loading testing mechanism two are separated or combined through a clutch, when the two are separated, single-channel loading testing and double-channel synchronous loading testing can be realized, and when the two are combined, double-channel coupled loading testing can be realized. The device solves the industry problem that the traditional ground testing is mismatched with the real flight working condition, and provides a reliable means for performance verification, fault reproduction, algorithm optimization and airworthiness certification of the flight control actuator system.

The present application relates to a kind of soil heavy metal with water migration quantitative prediction experimental equipment and method, belong to heavy metal with water migration experimental equipment field, including soil model preparation device, soil sample collection device and heavy metal detection device, soil model preparation device is by successively connected liquid storage barrel, first stirring barrel, drying barrel with heating function, second stirring barrel and soil forming barrel Composition, can complete soil and heavy metal liquid mixing, drying, secondary stirring and compaction forming, prepare the soil model with heavy metal;Soil sample collection device can be layered sampling to the multiple depth layer, multiple annular array sampling point around soil model leaching water place;Heavy metal detection device is made into target mixture by detecting box crushing, stirring soil sample, with heavy metal sensor detection sample heavy metal content, can realize the quantitative prediction experimental detection of soil heavy metal with water migration.

The invention relates to a circulating water tank earthquake simulation test system and test method based on electromagnetic driving and artificial intelligence simulation, a circulating water tank main body, a wave making system and a three-axis electromagnetic seismic station are integrated for the first time, X / Y / Z three-axis vibration can be realized, and each axis is independently controlled. A similar criterion library is constructed by combining a contact sensor and non-contact optical measurement, working condition errors of an actual environment and an experimental environment can be reduced as much as possible according to different experimental purposes, complex working conditions that a structure bears wave, water flow and earthquake effects at the same time in a real marine environment are simulated, and an experimental result has higher reference value. One-key process control is achieved through the central control unit, waveform generation and data processing are conducted through the AI technology, and the experiment efficiency and the scientific research depth are greatly improved.

The invention relates to the technical field of energy consumption management, and discloses an IBMS intelligent building energy consumption management system and method, and the system comprises a data collection unit which collects the internal and external environment and equipment operation information of a building; and the first processing unit constructs an operation feature set of each subsystem of the building based on the historical information, and obtains an operation evolution factor set. The second processing unit is used for acquiring an operation feature set and an operation evolution factor set of the to-be-detected building sample, comparing with the first processing unit, calculating an operation deviation and determining a deviation degree according to a preset grading standard; when the second deviation degree is detected, the third processing unit analyzes the operation characteristic change direction and the operation evolution factor set change trend of the to-be-detected building sample and judges the data accuracy; and the database management unit is used for storing a database containing the calibration sample, the operation feature set, the operation evolution factor set and the preset grading standard. The problems of data distortion and lack of dynamic verification can be effectively solved.