7results about How to "Small root mean square error" patented technology

This invention belongs to the field of medical signal processing technology, specifically a method and prediction system for constructing a three-dimensional tumor motion trajectory prediction model. It involves collecting temporal data of point clouds from the human chest and abdomen, as well as temporal location information of the tumor, and establishing a unified spatial coordinate system. The temporal data of the point clouds are divided into n time periods of total duration T1, and the temporal location information of the tumor is divided into n time periods of total duration T2, resulting in n training samples. The point clouds of the training samples are further divided into several pseudo-voxel sub-blocks according to the human projection plane, and K pseudo-voxel sub-blocks are selected based on motion saliency to obtain the temporal feature vector of the tumor motion. The temporal feature vector of the tumor three-dimensional motion trajectory prediction is used as input, and the corresponding tumor temporal location information is used as output to train an attention-enhanced dilated convolutional prediction network. This invention enables high-precision, non-invasive prediction of the three-dimensional tumor motion trajectory.

This invention provides a method and system for direction-of-arrival (DOA) estimation of minimum redundancy linear arrays based on the quantum giant trevally mechanism, belonging to the field of array signal processing. To address the problem of drastic performance degradation of minimum redundancy linear arrays under harsh conditions of impulse noise, this invention proposes using the mean square error of DOA estimation and the probability of successful estimation as core optimization indicators. This method effectively improves the performance degradation problem of arrays under impulse noise by constructing a mapping relationship between the positions of minimum redundancy linear array elements and DOA estimation performance, significantly enhancing the accuracy and robustness of the direction-finding system. The quantum swarm intelligence optimization method designed in this invention provides a promising solution path for the optimization problem through a global search strategy and fitness function design. This invention not only overcomes the applicability and performance bottlenecks of traditional methods under impulse noise environments but also has significant theoretical and practical application value for promoting the engineering application of array technology in high-precision direction-finding.

The invention provides an airfoil flow field sensor optimization arrangement method and system, and the method comprises the steps: obtaining the surface flow field parameter data of an airfoil under different working conditions, and constructing a data matrix; complementing the data matrix by using the matrix complementing model; and searching and optimizing the arrangement position of the sensor by taking the completion error of the matrix completion model as an optimization target to obtain an optimal sensor arrangement scheme. The method further comprises the steps that the reconstructed complete pressure distribution is compared with actual measurement data or historical simulation data of the partial redundant sensors, and if errors exceed a preset threshold value, re-optimization of the optimal sensor arrangement scheme is triggered. According to the method, the data detected by the sensor is recovered and extracted by using the matrix completion algorithm, so as to obtain information as much as possible; the genetic algorithm is used for optimizing the configuration position of the sensor so as to maximize the information carried by the sensor detection data; the problems of sensor efficient configuration and data reconstruction in flow field detection are solved.

The invention belongs to the technical field of deep sea navigation positioning, and provides an inertial navigation tight coupling correction method and system based on a time-sharing collaborative underwater acoustic beacon, and the technical scheme is as follows: in the navigation process of an AUV (Autonomous Underwater Vehicle), before acoustic auxiliary information arrives, a pure inertial navigation mode is adopted, and based on acquired IMU data and depth sensor data, inertial navigation calculation is carried out; obtaining an AUV navigation state containing an accumulative error; after the acoustic auxiliary information arrives, switching to a combined navigation mode for resolving, utilizing beacons at different positions to cooperatively broadcast acoustic information in a time-sharing manner, resolving the horizontal distance between the AUV and each beacon, and applying a vertical coordinate constraint; based on the measurement distance and the AUV navigation state containing the accumulative error, fusion processing is carried out through a tight coupling filtering algorithm, the accumulative error in the AUV navigation state is corrected, and a corrected AUV positioning result is obtained. And long-endurance and high-precision underwater positioning of the AUV is realized.

The invention relates to a method for predicting the cycle life of a mechanically damaged lithium battery based on mechanical characteristics and early cycle information, and belongs to the technical field of lithium ion batteries. The method comprises the following steps: firstly, performing defect manufacturing on the lithium battery under different indentation working conditions; then, finite element mechanical modeling is carried out on the lithium battery under different mechanical indentation working conditions; then carrying out a charge-discharge cycle test on the lithium battery containing mechanical damage to enable the lithium battery to reach the end of service life, obtaining cycle data of the battery, and carrying out electrochemical feature extraction on the cycle data; performing mechanical feature extraction on the lithium battery under different working conditions through a finite element model; and finally, predicting the service life of the defective battery under different indentation working conditions through the machine learning model. According to the method, the full life cycle prediction of the lithium battery can be carried out by using the early cycle data and the damage mechanical characteristics, valuable guidance is provided for the aging and service of the mechanically damaged lithium battery, and the development of a prediction tool is promoted, so that more effective battery management and performance optimization are realized.

This invention discloses a method and system for acquiring emitted longwave radiation based on infrared hyperspectral data. The method includes: acquiring infrared hyperspectral channel observation data and observation geometry information of a target satellite; filtering all effective channels based on channel signal-to-noise ratio and constructing an radiance matrix; performing principal component analysis on the matrix to extract the first n principal components; calculating initial radiation values ​​based on the principal components using a pre-established emitted longwave radiation inversion model; and interpolating the initial radiation values ​​calculated by multiple pre-established inversion models based on different observation angle intervals according to the satellite observation angle, ultimately obtaining high-precision emitted longwave radiation values. This invention fully utilizes the spectral information of all effective channels in the infrared hyperspectral system and corrects errors introduced by optical path differences through observation angle interpolation, significantly improving the accuracy and reliability of emitted longwave radiation inversion.

The application belongs to the technical field of ultrasonic nondestructive testing, and particularly relates to a multi-resolution sparse decomposition algorithm for ultrasonic signal separation and ultrasonic echo estimation in overlapping ultrasonic signal decomposition. The algorithm separates seriously overlapping ultrasonic echoes by continuously iterating the given ultrasonic signal, dividing the signal into two segments in each iteration, and decomposing each segment signal on a discrete Gabor dictionary with more refined discrete steps of parameters using a support matching pursuit algorithm. Meanwhile, a new discrete Gabor dictionary is generated by compressing the upper and lower boundaries of parameters in the Gabor function and refining the discrete steps of parameters in each iteration. Without increasing the number of atoms in the dictionary, the discrete interval of the dictionary parameters is gradually reduced to gradually improve the matching degree between the ultrasonic echoes and the obtained atoms after each algorithm iteration. The signal segments obtained by signal segmentation in each iteration are decomposed at a higher resolution than before, and the overlapping echoes that cannot be separated at a coarse scale can be separated at a higher scale, thereby gradually improving the accuracy of echo separation and echo estimation.