31results about How to "Guaranteed monotonicity" patented technology

A novel digital attenuator circuit and associated pre-power amplifier (PPA) that substantially increases the dynamic range of the amplifier. Increased dynamic range is achieved by placing a digital current attenuator circuit at the output of the pre-power amplifier so that the minimum possible current output of the transistor switch array of the PPA can be further attenuated. The attenuator functions to split the current between the load and the power supply V_DD (i.e. AC ground) based on device ratio that is controlled digitally via an input power control word. The digital attenuator is constructed as a segmented digitally controlled matrix or cell array comprising at least a pass and bypass matrix or array. The pass matrix controls the amount of current output from the PPA while the bypass matrix controls the amount of current shorted to the AC ground (i.e. power supply). By varying the number of transistors on or off in each matrix, the power output of the PPA can be easily and accurately controlled.

Provided is an integrated scheduling method for key loading and unloading resources of an automatic container terminal. The integrated scheduling method comprises the steps of setting chromosome coding modes, initializing population, defining the fitness function, setting the interlacing operation guideline and the mutation operation guideline, setting the parental generation selection strategy and the filial generation receiving strategy, using the fitness function to conduct chromosome iterative computation, and when meeting the iterative stopping criterion, stopping iteration and outputtingthe optimum solution. The integrated scheduling method can reasonably control device operation and space resource distribution of the container terminal, improves the operation efficiency of the automatic container terminal, reasonably distributes devices and space resources, avoids congestion, shortens the driving time of three types of devices, and optimizes space resource usage of a storage yard.

The invention relates to a detection restorative method of hydraulic valve jamming and chucking faults, which comprises a detection step, a judgment step and an execution step which are carried out through electric connection. The judgment step is that jamming and chucking membership degree judgment is given out through a fuzzy multi-criteria decision making theory method according to average value information of control signals output through the detection step, frequency of shimmy signals, peak-to-peak value and wave distortion. By means of the technical scheme, the detection restorative method can detect and judge the jamming and chucking faults of a slide valve plug in real time and repair the jamming and chucking faults caused by pollution and imbalance radial force by using supersonic vibration, and the detection restorative method is simple and convenient in method and low in cost.

The invention discloses a tumor prognosis prediction system based on a deep belief network. The system comprises a data collecting module, a data preprocessing module, a data learning prediction module and a prediction result display module, wherein the data collection module is used for collecting tumor information; the data preprocessing module is used for carrying out missing value processing and normalization processing on tumor original data; the data learning prediction module is used for carrying out deep learning and prediction modeling on the tumor data; and the prediction result display module is used for displaying a relative risk output by the data learning prediction module. A Gaussian restricted Boltzmann machine is used to keep the nonlinear characteristics of the data; according to the dimension of input data, the amount of output categories and the accuracy of a model, the deep belief network can be flexibly expanded; and in a model training process, no restriction and hypothesis is adopted, an influence way of a variable for the result and a mutual function among variables can be fully mined, the influence way of different factors for tumor prognosis can be comprehensively revealed, and tumor prognosis prediction accuracy is improved.

The invention relates to a control method and a device of a power converter of a resonance topological circuit, wherein the control method of the power converter of the topological circuit comprises the steps of regulating the frequency of the power converter and the combination of the duty ratio to ensure that the power converter has good monotonicity in the whole output range of the control device. The control device of the power converter of the resonance topological circuit comprises a control module, which is used for regulating the frequency of the power converter and the combination ofthe duty ratio to ensure that the power converter has good monotonicity in the whole output range of the control device. When being implemented, the method and the device of the invention can ensure the monotonicity of the power converter of the resonance topological circuit in the whole PWM work range.

The invention relates to an intelligent jamming and chucking detection controller for a hydraulic valve. The intelligent jamming and chucking detection controller is electrically connected with an electro-hydraulic proportional valve provided with a proportional electromagnet and a valve element, and comprises a flutter signal generation module (16), a system control signal module (14), a stabilized power module (15), a proportion integration differentiation (PID) control module (17), a power amplification module (18), a sampling module (19), a signal processing module (20), a negative feedback module (21) and a jamming and chucking decision-making module (22), wherein the jamming and chucking decision-making module (22) makes jamming and chucking membership decision by a fuzzy multi-criteria decision-making method. By the controller, a jamming and chucking failure of a valve element of a sliding valve can be detected and judged in real time; and the controller is high in sensitivity and accuracy, simple in structure and low in cost.

The invention belongs to the technical field of side-scan sonar image processing, and particularly relates to a nonlinear enhancement method for a side-scan sonar image. The method comprises the following steps of: preparing a substrate; obtaining an original image from the original side-scan sonar data file; sequentially performing gray normalization, neighborhood extremum suppression and Gaussian smoothing on the original image, calculating a maximum value of a low-gray-scale region, a minimum value of a high-gray-scale region and an intermediate gray-scale region for the smoothed image, andfinally respectively performing nonlinear correction on the three regions to obtain an enhanced image. Through the steps, enhancement of effective signals and suppression of invalid signals of the original side-scan sonar image can be rapidly and effectively achieved at low cost, the contrast ratio of local features of the image is enhanced, monotonicity of edge and gray level distribution of theoriginal image is kept, and false edges and the like are not introduced.

The invention provides a data mining method capable of guaranteeing quality monotony. The method comprises the following steps: after an original big data set is compressed by a PCA (principal components analysis) technology, mapping the original big data set onto an R-tree data structure; then, carrying out mining processing on the data set by an improved K-nearest neighbor classification algorithm. The method mainly comprises the following two parts including a coding part and a mining part, wherein the coding part utilizes R-tree to present data, data with high similarity in the data is combined to serve as one node of the R-tree so as to achieve a purpose of mass data compression and improve the efficiency of the mining part; the mining part utilizes the thought of the improved K-nearest neighbor classification algorithm to process the data node and predict the classification of an input test point. According to the large-scale data mining method, the problem that the quality of a mining result and resource restriction cannot be balanced and the quality monotony of an approximate result cannot be guaranteed when big data is mined by a traditional algorithm under the restriction of limited time and resource restriction can be solved.

A programmable gain amplifier with three stages uses fine steps, has a large gain range, and is monotonic. The first stage comprises several amplifiers, each including a resistive feedback loop. The feedback loop comprises a series of resistors, with each resistor acting as a tap. Since the number of resistors in the loop is unchanging, monotonicity and stability is guaranteed when resistance is increased using successive taps. A switch system connects two taps at a time to an interpolation stage. Each of these taps corresponds to a specific resistor level, and thus a gain level. The interpolation stage uses a plurality of current sources inside a feedback amplifier to control the interpolation, in order to provide fine gain steps.

The invention discloses a spacecraft-folding-unfolding-structural-health-monitoring-oriented sensor configuration optimization method. The method includes the steps that (1) the alternative sensor data number n is determined, and the final-configuration sensor number m and the sampled mode order N are determined; (2) a structural dynamics-characteristic-equation solving equation is established to obtain a modal matrix; (3) the position of a finite element node serves as an optimization variable d; (4) a fitness function f<1> (d) based on a Fisher information matrix is established; a fitness function f<2> (d) considering the sensor configuration distance uniformity and significant intervals is established; the f<1> (d) and the f<2> (d) are optimized respectively with function maximization as an optimized target, and an optimal fitness function value f<1><*> and a value (The value is defined in the description) are obtained respectively; (5) the f<1><*> and the value (The value is defined in the description) are used in cooperation with the putting reliability degree of sensors on corresponding nodes, and a last fitness function f<4> (d) is established, wherein the f<4> (d) is optimized with function maximization as an optimized target, an optimal fitness function value (The value is defined in the description) and a corresponding optimization variable value (The value is defined in the description) are obtained, and the optimization variable value (The value is defined in the description) is a final sensor configuration position.

The invention provides an optimal binning data processing method and system based on an NSGA-II genetic algorithm, and the method comprises the steps: carrying out the preprocessing of a data sample, and obtaining a first data sample; performing pre-binning on the first data sample according to a pre-binning rule to obtain n pre-binning boxes; according to the n pre-separation boxes, defining a decision variable matrix which comprises a lower triangular matrix with the size of n, obtaining an IV value of the decision variable matrix; defining a target function vector according to the decision variable matrix and the IV value; setting constraint conditions of box separation; performing multi-objective optimization solution on the first data sample after pre-binning by using an NSGA-II genetic algorithm according to a binning constraint condition and an objective function vector to obtain a plurality of optimal solutions; obtaining an optimal segmentation point according to the plurality of optimal solutions; and according to the optimal segmentation point, carrying out binning on the first data sample after pre-binning.

The invention provides a pumping unit indicator diagram acquisition method based on polynomial fitting. The pumping unit indicator diagram acquisition method comprises the following steps: a wellheadcontroller RTU acquires N point set load displacement original data at fixed intervals according to a stroke period, and recombines indicator diagram data (Si, Fi) of dead points below the original data; data enhancement is carried out on the recombined displacement data points, and the data points are expanded to (1 + k) N point data sets Snew; n-order polynomial fitting is performed on the data,and polynomial coefficients a0, a1... an are calculated to obtain a Y (xi) polynomial function; and xi substitution calculation is carried out on the obtained function, so as to obtain a new displacement data set, and enhanced data is utilized and removed, and then new N-point indicator diagram data is formed. Through the polynomial curve fitting method, the pumping unit indicator diagram acquisition method is simple and efficient, and the stability and accuracy of displacement data acquisition are well improved.