55results about How to "Mathematical model is simple" patented technology

The invention discloses a method for analyzing crowd density base on statistical characteristics. The method comprises the following steps: video input and frame extraction are carried out; the mosaic image difference MID characteristics are extracted from the video-frequency frame sequence, and the subtle movements in the crowd are detected; the uniform distribution of the sequence time of the mosaic image difference MID characteristics is checked; the geometric correction is performed to the crowd and the scene with obvious perspective phenomenon, and a contribution factor of each picture element to the crowd density is obtained on the image plane; and the weighting process is performed to the crowd space area, so as to obtain the crowd density. Compared with the prior method, the method has no need of the reference background, also has no need of the background modeling and can self-adapt change of either morning or evening light, the algorithm is quite robust, and the application is convenient; the mathematical model is simple and effective, the spatial distribution and the size of the crowd can be accurately located, and the vivacity is strong; the calculation amount is small, and the method is suitable for real-time visual monitoring. The invention can be widely applied to the monitoring and the management of the public places with detained crowd density such as the public transportation, the subway, the square and the like.

The invention provides a large power grid construction economic evaluation method, which comprises the following steps that: a two-stage evaluation system is built; proper evaluation indexes are selected; the evaluation is represented by comments in different degrees by fuzzy languages in fuzzy comprehensive evaluation; then, factors which are not easy to quantify are quantified by the fuzzy relationship synthesis principle; and an evaluated object is comprehensively evaluated from a plurality of factors. Particularly, evaluation indexes are firstly determined; then, the membership degree and the weight of the indexes are calculated; a fuzzy evaluation matrix and a weight vector are subjected to synthesis calculation; and finally, a fuzzy comprehensive evaluation result is obtained. The evaluation result which is not easy to quantify is quantified, and the evaluation result is simple and visual. The method provided by the invention has the advantages that a mathematic model is simple; the calculation is simple and convenient; the evaluation effect on multi-factor and multilayer complicated problems is better; and the method is applicable to the large power grid construction economic evaluation study.

The invention discloses a pose synchronization method and device for an arc locus motion of a robot. The method includes the steps of acquiring pose data, calculating a quaternion number, establishinga quaternion pose spline curve, performing velocity planning, obtaining each interpolation node, calculating a quaternion pose and an Euler angle pose, and calculating a synthetic pose. The device includes a memory for storing a program and a processor for loading a program to perform the method of the invention. The method and device can solve the problem that it is difficult to collect the center of the circular arc of the robot, and achieve the better approaching to a machining curved surface, so that the posture change of the robot reaches the second-order continuous state, achieves the synchronization of the posture at an intermediate point, achieve the conversion of the planning of the posture into the planning of posture nodes, avoids the tedious process of integrating the pose into angles and then performing planning and re-interpolating to form the nodes, and avoids the condition that the physical meaning of each variable is not clear when the physical quantities are converted. The method and device can be widely used for the technical field of automatic control of robots.

The invention discloses an electroencephalogram signal unmanned platform intelligent control method based on a deep convolutional adversarial network. The method comprises the steps that a terminal carries out the noise removal of a collected electroencephalogram signal, and obtains a denoised electroencephalogram signal; performing deep feature extraction on the denoised electroencephalogram signal through a capsule network to obtain a deep feature signal; fusing the deep feature signal and the electroencephalogram signal and then carrying out classification and recognition to determine a corresponding control instruction signal; and the terminal performs offline and online test verification on the unmanned platform, and after verification succeeds, the unmanned platform receives and executes the control instruction signal sent by the terminal. According to the method, the existing noise data are integrated into the one-dimensional electroencephalogram signal training network, the mathematical model is simplified, the problem of insufficient noise training data is solved, the one-dimensional prediction signal is reconstructed by using the auto-encoder architecture, the attention mechanism is used for feature selection, and the calculation efficiency is improved.

The invention discloses a device and a method for calculating the coordinate of a track and the coordinate of a detected point. The method is characterized in that a high-precision CPIV track datum net point is utilized as the measuring basis of a three-dimensional detection system of the track; a benchmark total station and an auxiliary total station directly capture a CPIV datum net prism in order to obtain coordinate values of the CPIV datum net prism on two total stations; by calculating position difference between the benchmark total station and the auxiliary total station, influence of connection error for left and right car frames upon positions of total stations is eliminated; and the coordinate of the detected point of the track and the coordinate of a center line of the track are calculated under the principle of euler angles and the law of rigid body motion. A track detection trolley comprises a right trolley frame, the auxiliary total station, the benchmark total station, a left trolley frame and a dual tilt sensor. The method for the coordinate of the detected point and the track detection trolley have following beneficial effects: a digital model of a detection system is simplified via principles of euler angles and the law of rigid body motion; the coordinate of the detected point of the track and the coordinate of the center line of the track can be rapidly calculated; and detection precision and efficiency are enhanced.

The invention discloses a seafloor shallow stratum sound energy attenuation model building and two important sound energy attenuation feature parameter extraction method. The method mainly comprises the following steps: (1) according to a sonar equation, the sound energy level of reflected waves from the seafloor received by a receiving transducer is calculated; (2) starting from the seafloor, the echo sound energy level of each layers of seafloor shallow stratum sediments is calculated sequentially; (3) according to a surveying adjustment theory, the mathematical model for the whole sound energy propagation process of the shallow stratum is deduced; and (4) according to a slow change rule of sediment types of the shallow stratum in a transverse small range and further according to a least square rule, the back scattering difference and the attenuation compensation residual of neighboring traces on the same stratum are approximately equal. The attenuation rule of sound waves in the seafloor shallow stratum sediments is disclosed, two important acoustic parameters capable of reflecting physical features of the sediments are deduced, important acoustic feature parameters can be provided for classification of shallow stratumsediment types, and the sediment classification precision is better improved.

The invention discloses an improved PCNN compensation-based AUV image fusion method, a processor and a system. The method comprises the following steps of subjecting registered AUV source images to NSCT decomposition respectively so as to obtain a corresponding low-frequency sub-band coefficient and a corresponding high-frequency sub-band coefficient; for the low-frequency sub-band coefficient, selecting a low-frequency sub-band fusion coefficient by adopting the rule of improving the spatial frequency; selecting a high-frequency sub-band fusion coefficient for two layers largest in the direction number of the high-frequency sub-band coefficient by adopting a rule largest in absolute value, and selecting a high-frequency sub-band fusion coefficient for all the rest layers by adopting a self-adaptive PCNN model; performing NSCT reconstruction on the selected low-frequency sub-band fusion coefficient and the selected high-frequency sub-band fusion coefficient, and finally obtaining a fused image. The fusion quality of images is effectively improved.

The invention discloses a hot-rolled steel plate residual stress calculating method suitable to production field, belonging to the technique field of the hot-rolled steel plate residual stress calculating method. The calculating method is realized by the steps of sampling straightened hot-rolled strip steel and cutting the sample to be narrow strip steel, wherein a calculating formula is Sigma R=8*Delta x*E*d1/L<2>, the measurement of parameters delta x, d1 and L is simple in the operating method and low in technical requirements, and E is determined by the self-owned properties of the strip steel. The calculating method has a simple mathematical model and intuitive and reliable in the calculating result.

The invention discloses a virtual human eye emotion expression simulation method, which comprises the following steps of: firstly, acquiring a three-dimensional human eye mesh model, namely, three-dimensional mesh data; secondly, selecting mesh points at eye areas of the model according to the motion characteristics of eyes as motion control points; and finally, driving the motion of boundary control points by adopting a crank and rocker mechanism model to simulate real eye motion. In the method, the same mathematic model is adopted to realize the repeated motion process of opening and closing the eyes under the condition of continuous time and produce richer eye closing motion reflecting different emotions by regulating parameters such as position relationships among different rods, initial mounting position of a crank and the like. In the method, the mathematic model is simple; and the method is explicit in physical significant, simple to control, easy to implement and high in reality.

The present invention discloses method of measuring atomic force microscope needle tip radius with single-wall carbon nanotube. The measurement includes the following steps: depositing single-wall carbon nanotube of 1-5 nm diameter on the surface of mica chip; measuring with atomic force microscope with the needle tip to be measured the measurement width w and height h of single-wall carbon nanotube; and calculating the radius of the needle tip to be measured based on w and h. The present invention uses the single-wall carbon nanotube as standard measuring sample for measuring the radius of the needle tip of atomic force microscope, and has the advantages of simple and intuitive mathematics mold clear physical meaning, precise, efficient and real-time measurement, etc. and may be used in measuring needle tip radius over 1 nm.

A method for controlling an ECM motor to output a constant torque. The method includes: 1) entering a target torque value T0 from an external device; when the motor is in a non-use state, starting the motor, and allowing the microprocessor to acquire an original output voltage value P chopped by a PWM signal; when the motor is in a running state, allowing the microprocessor to acquire a current output voltage value P chopped by the PWM signal; 2) enabling the microprocessor to calculate a target bus current value Itad using the function Itad=F(T,P) according to the target torque value T0 and the output voltage value P chopped by the PWM signal; and detecting a real-time bus current Ibus; and 3) allowing the microprocessor to compare the target bus current value Itad with the real-time bus current Ibus for conducting a closed-loop control.

The invention discloses an online detection device of the clear water layer thickness of a concentration tank, and aims at solving the problems of lag control and poor precision of the existing device. The online detection device is characterized by comprising a detection rod, a pressure sensor, a guide base, a supporting base, a lifting device, a displacement sensor, a programmable logic controller (PLC) control box, an upper travel switch and a lower travel switch, wherein the supporting base and the guide base are arranged on the bridge frame of a concentrator; the lifting device and the upper travel switch are arranged on the supporting base, and the lower travel switch is arranged on the guide base; the displacement sensor is fixed on the side face of the lifting device, and the lower side is connected with the detection rod; the pressure sensor is arranged in the inner cavity of the detection rod, and a probe stretches out and contacts with the level; and a limiting plate is arranged on the detection rod, and the upper plane of the limiting plate is constantly contacted with the contact of the displacement sensor. According to the online detection device of the clear water layer thickness of the concentration tank, by the combination of the pressure sensor and the displacement senor, and by an automatic conversion system and the guide device which move up and down, the thickness of the clear water layer in the concentration tank can be automatically detected in real time, so that the automatic control of the concentration process is achieved, and the labor intensity of the operator is reduced.

A method for measuring an air volume of a blower motor, including: a) establishing M constant air volume control functions Qi=F(n) corresponding to M air volume points CFM_i in a microprocessor of a motor controller; b) starting the motor, measuring a real-time rotational speed n₀ and a parameter Q₀ of the motor after the motor runs in a stable state, putting the parameter Q₀ into the functions Qi=F(n) to obtain multiple rotational speeds n_i, comparing the real-time rotational speed n₀ with the multiple rotational speeds n_i corresponding to the multiple air volume points CFM_i, and determining a real-time detected air volume CFM₀ which falls within two air volumes CFM_i and CFM_i-1; and c) calculating the real-time detected air volume CFM₀ according to the two air volumes CFM_i and CFM_i-1 and the rotational speeds n_i and n_i-1 using an interpolation method.

The invention discloses an ECM (electrochemical machining) motor and method for constant torque control based on an accurate flux linkage look-up table method, and relates to the field of motors. As busbar current easily fluctuates, control accuracy is reduced. The ECM motor comprises a motor body and a controller for controlling work of the motor body, the motor body is a magnetic steel insertiontype motor with a reluctance torque, a shell of the motor and the controller are integrated or split, the controller is provided with a storage module, the storage module is used for storing a d-axisflux linkage motor parameter list of a relationship among a flux linkage lambda d, Id and Iq and a q-axis flux linkage motor parameter list of a relationship among a flux linkage lambda q, Id and Iq,the controller can conveniently and accurately control a constant torque through motor parameter list data in the storage module and the look-up table method, busbar current sampling is omitted, control is more accurate, and cost is reduced.

A method for calibrating and controlling parameters of a motor includes the steps of calibration and motor control. The method includes the following steps that through an experiment, complex control parameters required by motor control are linearized into a corresponding control parameter data table using torque and speed as conditions, a current sensor and the like are used for correcting performance deviation caused by the influence of other conditions, accurate and efficient motor control is achieved, a mathematical model required by the motor control is simplified, limitations of the mathematical model on motor control are overcome, and efficient and accurate control over the motor is achieved.

The invention discloses a method for estimating kinetic parameters of a spatial non-cooperative target under a unilateral constraint condition. The method comprises the following steps of regarding the relative motion of a target and the tail end of a mechanical arm as the newly-added freedom degree of a Lagrange kinetic system, on the basis of the assumption, using an extended freedom degree model for describing the relative motion between the mechanical arm and the target, and establishing a kinetic model; analyzing the combination kinetic matrix line by line, separating different parametersaccording to parameter correlation, and writing a kinetic model into a compact form about unknown parameters; and iterating the guessed value error to construct a least square model so as to realizemeasurement and extraction of non-cooperative target kinetic parameters. The defects that when an existing mechanical arm system grabs an unknown target, whether relative sliding occurs or not cannotbe detected, and target kinetic parameters cannot be accurately measured under the sliding condition are overcome.

A method for acquiring a constant torque of an ECM, the method including: A) acquiring a target torque value T0 input from external; B) when the motor is in a non-use state, operating the motor and acquiring an initial rotational speed rpm by the microprocessor; and when the motor is in an operating state, acquiring a current rotational speed rpm by the microprocessor; C) calculating a corresponding target bus current value Itad by the microprocessor using the function of a DC bus current Itad=F(T, rpm) according to the target torque T0 and acquired rotational speed rpm, in which T represents a torque value output by the motor; and D) comparing the target bus current Itad with a real-time bus current Ibus by the microprocessor in a closed-loop control according to the detected real-time bus current Ibus.

The invention discloses an Internet of Things photovoltaic power generation control system and method based on GA-SVM modulation, and the method comprises the steps: carrying out the parameter setting of a first current modulation module, a second current modulation module, a PR controller and a notch filter, and receiving the voltage and current of a common point of a smart power grid, the three-phase current of an output end of an inverter, and the DC output voltage of a photovoltaic cell; acquiring and judging whether to generate a first reference current and/or a second reference current in real time, and determining one of the first current modulation module and the second current modulation module with the optimal response speed; if the response speed of the second current modulation module is optimal, corresponding switching is carried out, so that a connecting line between the second current modulation module and the SVM wave sending module is switched on, and voltage Va, Vb and V0 are obtained through conversion of second reference current; otherwise, converting the first reference current to obtain voltages Va, Vb and V0; and based on SVM wave generation, the inverter is controlled to work, the response speed is high, and the power generation efficiency is high.

The invention discloses a three-phase nine-switch dual-synchronous motor time-sharing control inverter and a control method thereof. A direct torque method is used to obtain a flux linkage, a torque and a flux linkage angle of a dual-permanent-magnet synchronous motor, switching state vector division sectors of two three-phase permanent magnet synchronous motors are obtained, four directions of torque increase, torque reduction, flux linkage increase and flux linkage decrease can be divided according to sector positions and according to motor rotors of different times, thus bridge arm switch on and off states in different directions can be obtained, the voltage vector tables of the three-phase permanent magnet synchronous motors are obtained by direct torque control, the switching states of the bridge arms are selected through a weight function to control the two three-phase permanent magnet synchronous motors, selection switches can independently control the input of PWM signals of different odd and even times into the inverter, and the time-sharing operation of the two three-phase permanent magnet synchronous motors is achieved. According to the method, the control is simple andclear.

The invention provides a real-time and accurate target recognition system based on machine vision and a corresponding tractor reversing and farm tool connection method. The method comprises the steps:accurately extracting a special target pattern or marker of a feature point on a farm tool through deployment, which are used for real-time detection and relative position calculation of a target; substituting the feature points of the target pattern in the camera imager into a corrected camera model, so the position of the target relative to the camera can be accurately derived, i.e., the distance and a direction angle. Due to the fact that the camera and the tractor are relatively static, the relative position and direction information of respective connecting points of the agricultural implement and the tractor can be accurately measured, a navigation path is calculated and planned, and the suggested vehicle speed and the steering wheel angle are provided for a driver to refer to so that the driver can manually or automatically drive the tractor to reach the agricultural implement mounting position. Compared with a camera for directly detecting a machine, the method can greatly reduce the calculation amount of detection, and improves the accuracy, reliability and calculation speed.