31results about How to "Avoid loss of precision" patented technology

The invention provides a method and a device for realizing an image container control with the effect of a convex lens. The method comprises the following steps of: collecting the offset delta x of x-coordinate generated by moving figures on a touch screen at intervals per unit time using the top left corner of a main window as the origin of coordinate of the main window; calculating the X coordinate Xw(i) of the origin of coordinate of a display window according to the offset of the coordinate x generated by moving figures at intervals per unit time and calculating the zoom coefficient according to the X coordinate XC in the centre point of the display window and the X coordinate X(i) in the centre point of the subcontrol; obtaining the width W(i) of a new image and the height H(i) of a new image according to the zoom coefficient; calculating coordinate X at the top left corner of the new image according to the X coordinate X(i) in the centre point of the subcontrol, the width W(i) of the new image and the X coordinate Xw(i) of the origin of coordinate of the display window; calculating the coordinate Y at the top left corner of the new image according to the Y coordinate H in the centre point of the subcontrol and the height H(i) of the new image; and drawing the image according to the parameters. The invention realizes the image container control with the effect of the convex lens in a way of light-weight resource consumption in case of limited CPU and memory resources.

The invention relates to a machinery device for improving resolution of an angle encoder and measurement precision. The device comprises a connection coding disc rotating shaft bearing seat (1), connection coding disc rotating shaft bearings (3), a connection coding disc rotating shaft (4), a coding disc (5), a driven gear (6), an angle coder base seat (7), input shaft bearings (8), an input shaft (9) and a driving gear (10), wherein the input shaft (9) is mounted on the angle coder base seat (7) through the two input shaft bearings (8), the connection coding disc rotating shaft (4) is mounted on the connection coding disc rotating shaft bearing seat (1) through the two connection coding disc rotating shaft bearings (3), the connection coding disc rotating shaft bearing seat (1) is mounted on the angle coder base seat (7), the driving gear (10) is arranged on the input shaft (9), the driven gear (6) is arranged on the connection coding disc rotating shaft (4), the driven gear (6) and the driving gear (10) are engaged, the input shaft (9) is connected with a detected object, and the coding disc (5) is arranged on the connection coding disc rotating shaft (4). The machinery device greatly improves the resolution and measurement precision by adding a variable gear transmission device.

The invention relates to the technical field of dynamics system optimization, and provides a flexible multi-body system dynamics semi-analytical sensitivity analysis method based on absolute node coordinate description. The method comprises the following steps: firstly, establishing a mass matrix, a rigidity matrix and a generalized force array of the flexible multi-body system based on an absolute node coordinate method; secondly, establishing a kinetic equation and an optimization objective function of the flexible multi-body system; thirdly, based on a direct differential method or an adjoint variable method, establishing a semi-analytical sensitivity calculation formula of the flexible multi-body system dynamics; and finally, solving the dynamic differential algebraic equation of the flexible multi-body system to obtain a sensitivity calculation result. The method is based on an absolute node coordinate method and a multi-body system dynamics theory. A semi-analytical sensitivity calculation formula of the flexible multi-body system is established to solve the sensitivity analysis problem of dynamics of the flexible multi-body system, a set of new strategy for sensitivity analysis of the flexible multi-body system is provided, and convenience is provided for sensitivity calculation of the large-scale complex flexible multi-body system.

The invention discloses an image sub-pixel matching and positioning method and system, equipment and a medium. The image sub-pixel matching and positioning method comprises the steps: preprocessing a preset template image and a target image; carrying out pixel-level similarity rough matching on the pre-processed preset template image and the target image, and obtaining rough matching translation coordinates and a rough matching rotation angle; taking the coarse matching translation coordinates and the coarse matching rotation angle as initial parameters of a cost function in a preset similarity matching model, and determining a loss value corresponding to the similarity of the sub-pixel level between the preset template image and the target image through the cost function with the initial parameters; when the loss value meets a convergence condition, taking the initial parameter as a final positioning result parameter of the target image; and performing preset positioning operation on the preset positioning object according to the final positioning result parameter. According to the method and system, the problem of accuracy loss caused by discretization of translation coordinates and rotation angles can be effectively avoided, and the positioning accuracy is improved.

The invention provides a networked automobile battery management system. A networked automobile comprises an automobile battery for providing a power source for the networked automobile, and the networked automobile battery management system comprises a cloud server and a vehicle-mounted controller located at a vehicle-mounted end. Based on the networked automobile battery management system, the invention further provides an SOC and SOH combined estimation algorithm. The SOC and SOH combined estimation is carried out in a cloud server, so that the algorithm precision loss caused by fixed-pointnumber operation and index lookup approximation of the vehicle-mounted controller is avoided, and the SOC and SOH combined estimation algorithm based on cloud computing can improve the SOC estimationprecision by 2.5-4.5%. The SOH estimation algorithm provided by the invention not only can reflect the DC internal resistance of the battery, but also can calculate the aging state of the internal resistance of the battery under different frequency conditions in real time, characterizes the aging trend of the polarization characteristics of the battery, and can reflect the SOH of the battery morecomprehensively.

The invention relates to the technical field of a boat analog simulation technology, and provides a boat body surface meshing method and device for hydrodynamic calculation. The method comprises the following steps of 1, reading a boat body STL model to obtain the boat body scale information; 2, performing slicing processing on the boat body STL model by using the obtained boat body scale information to obtain all slice data of the boat body STL model; and 3, using all slice data of the boat body STL model to divide boat body surface lattices. The boat body surface lattices in the boat body hydrodynamic calculation can be fast generated; the generation process is simplified; the dense and precise boat body sectional line data can be accurately obtained; the errors due to three-dimensionalinterpolation and curve-fitting can be avoided; the real-time performance of the algorithm is good; and various complicated boat types difficult to be processed by the existing algorithm can be processed.

The invention discloses a Bayesian regularization back propagation neural network coordinate conversion method and device, which are used for solving the problems that in the coordinate conversion process, a set of coordinate conversion parameters is adopted in the whole area, so that the precision of the local area is lost in the coordinate conversion process, and the accuracy of the converted coordinates is influenced. The Bayesian regularization back propagation neural network coordinate conversion method comprises the following steps: determining plane coordinates of a 2000 country geodetic coordinate system and plane coordinates of an engineering independent coordinate system of a control point, and creating a learning set; training a neural network model according to the learning setuntil the performance index of the neural network model reaches a preset value, and determining that the training is completed; and according to the trained neural network model and the 2000 countrygeodetic coordinate system plane coordinates of the to-be-measured point, determining engineering independent coordinate system plane coordinates of the to-be-measured point obtained through conversion. According to the Bayesian regularization back propagation neural network coordinate conversion method, the network weight of the back propagation neural network is limited through the Bayesian regularization algorithm, and the network structure is effectively improved, and the accuracy of coordinate conversion is improved.

The invention relates to a drilling and marking device. The drilling and scribing device comprises a base; the first sliding assembly comprises a first sliding rail and a first sliding block, the first sliding rail is arranged on the top face of the base, and the first sliding block is arranged on the first sliding rail in a sliding mode; the second sliding assembly comprises a second sliding rail and a second sliding block, the second sliding rail is arranged on the top face of the base, the second sliding block is arranged on the second sliding rail in a sliding mode, and the first sliding rail is perpendicular to the second sliding rail; the measuring unit is used for measuring the sliding distance of the first sliding block and the sliding distance of the second sliding block; a first slide bar and a second slide bar; the supporting seat comprises a first through hole and a second through hole, the supporting seat is arranged on the first sliding rod in a sliding mode through the first through hole, and the supporting seat is arranged on the second sliding rod in a sliding mode through the second through hole; the first scribing needle is arranged on the supporting seat, and the axis of the first scribing needle is perpendicular to the top surface of the base. According to the drilling lineation device, the lineation and drilling efficiency and precision are improved.

The invention relates to a matrix inversion method, device and equipment for a multi-antenna system and a storage medium, and relates to the technical field of wireless communication, and the method comprises the steps: obtaining an interference noise covariance matrix in the multi-antenna system, wherein the interference noise covariance matrix is a four-order square matrix determined based on noise sequences of four receiving antennas in the multi-antenna system, and two elements symmetrical along a main diagonal in the interference noise covariance matrix are conjugated and equal; decomposing the interference noise covariance matrix into four sub-matrixes, wherein each sub-matrix is a second-order square matrix; and performing inversion calculation on the interference noise covariance matrix based on the four sub-matrixes, and determining an inversion result of the interference noise covariance matrix. In this way, decomposition and order reduction can be carried out on the four-order interference noise covariance matrix, and the inversion result can be directly obtained by omitting the layer-by-layer iteration process, so that the precision loss in the iteration process can be avoided, the bit error rate in the multi-antenna system is reduced, and meanwhile, the operand is reduced.

The invention relates to the technical field of electric screwdrivers, and provides an automatic intelligent electric screwdriver based on a torque and pressure composite force sensor and a control method of the automatic intelligent electric screwdriver. The automatic intelligent electric screwdriver based on the torque and pressure composite force sensor comprises a base; a motor; one end of the torque and pressure composite force sensor is fixed on the base, and the other end of the torque and pressure composite force sensor is connected with the motor; the screwdriver head structure comprises an electric screwdriver core shaft and a screwdriver head, and the two ends of the electric screwdriver core shaft are coaxially connected with the motor shaft and the screwdriver head respectively; and a sensor aviation plug and a motor aviation plug are arranged on the fuselage shell. According to the automatic intelligent electric screwdriver based on the torque and pressure composite force sensor, tightening torque and assembly pressure data in the screw locking process can be collected and monitored in real time, and the consistency and reliability of operation quality and the controllability of assembly pressure are greatly improved while ultrahigh tightening precision is guaranteed.

The embodiment of the invention provides a normalization quantification method and device, electronic equipment and a storage medium. The method comprises the following steps: obtaining quantized data Qconv serving as input data of a normalization module; calculating a parameter mean value Q [mu] and a standard deviation Q [sigma] required by normalization of the quantized data Qconv; (Qconv-Qmu) * IntA is calculated, SR is calculated again according to a calculation result of the (Qconv-Qmu) * IntA, and the SR is determined according to a re-scaling parameter of the normalization module and a parameter of a subsequent module of the normalization module; and determining a full-fixed-point reasoning result of the neural network according to the result. According to the normalization quantification method provided by the embodiment of the invention, the problem of precision loss caused by normalization in a fixed-point calculation process is avoided by utilizing a thought of first calculation of (Qconv-Qmu) * IntA, so that full fixed-point reasoning of a neural network model with an INGNLN module becomes possible, the reasoning speed of the neural network model is improved, and the reasoning efficiency of the neural network model is improved. And the power consumption of model reasoning is effectively reduced.