169results about How to "Improve speed measurement accuracy" patented technology

A wireless sensor network system using GMR magnetic sensitive technology to detect vehicle information of the invention consists of two GMR sensor detecting end machines and a wireless receiving end machine. The two GMR sensor detecting end machines are connected with the wireless receiving end machine. The detection method of the invention detects information such as vehicle speed, vehicle length, driving direction, vehicle flow, lane occupancy ratio, etc., according to change data of the surrounding ground magnetic field sensed by the GMR sensor when the vehicles pass. The wireless sensor network system is characterized in having high precision of detecting speed, adaptability to various roads, low cost, convenient maintenance and high anti-interference ability and causing small destroy to road surface, etc.

The invention discloses a velocity measuring method of a ground motion object by use of an unmanned plane. The method comprises the following steps: an aviation photoelectric surveillance platform carries out object searching in a relatively quite large scope, after the position of a search area is given, a visible light camera is driven to rotate to enable an object to enter a tracking view field, and the view filed is searched by use of a stepping and staring combined mode; after a motion object image of interest is obtained, an image tracker starts working, isolates aerial carrier attitude change and interference of low-frequency vibration, enables the object to be always maintained at the center of the view field of the camera and realizes stable tracking of the object; it is maintained that the aviation photoelectric surveillance platform stably tracks the motion object, and coordinates of the motion object at each time point under a reference coordinate system are solved every equal time period; and calculated coordinate data of multiple objects under the reference coordinate system are substituted into a parameter equation of an object motion locus, unknown parameters in the reference equation are fitted by use of a least square method, the derivative of the reference equation for time t is respectively solved once and twice, and an object speed and an accelerated speed are solved.

The invention discloses a vibration measuring method of engine speed by applying a four-point energy centrobaric correction method, which comprises the steps of: discrete-sampling a vibration acceleration signal on the surface of an engine block to obtain a time series; obtaining a frequency domain signal X1(f) through fast Fourier transformation of Zn(t); correcting the peak frequency fi of X1(f) by adopting a four-point energy centrobaric correction method, and obtaining a corrected frequency; obtaining a time spectrum signal ZZn(t) through carrying out self-related transformation and fast Fourier transformation on the frequency domain signal X1(f); correcting the time ti corresponding to the peak of ZZn(t) by adopting the four-point energy centrobaric correction method again, and obtaining the corrected time; and finding the accurate 1 / 2 harmonic frequency value of the rotating speed w of the four stroke engine according to the corrected frequency and time, and calculating the rotating speed w of the engine through a relational expression. The vibration measuring method of engine speed can solve the problem that the rotating speed of the engine cannot be detected when the number of engine blocks cannot be determined; and the method has the advantages of better universality, high noise-proof performance and accurate and reliable result.

The invention belongs to the rail transit technical field and provides a train speed measuring method, a processor and a train speed measuring system. The method includes the following steps that: the speed measuring estimated value of a train at the current period is calculated based on the overall speed measuring value of the train at a last period and the acceleration of the train at the last period; the acceleration of the train is acquired by an acceleration sensor group installed on the train; the speed measuring values of radars of a radar group installed on the train at the current period and the speed measuring values of cylindrical grating speed sensors of a cylindrical grating speed sensor group installed on the train at the current period are obtained; the error estimated values of the speed measuring values of the radars at the current period and the error estimated values of the speed measuring values of the cylindrical grating speed sensors at the current period are calculated based on the speed measuring estimated value of the train at the current period; and the overall speed measuring value of the train at the current period is calculated according to the error estimated values of the speed measuring values of the radars and the error estimated values of the speed measuring values of the cylindrical grating speed sensors. With the train speed measuring method, the processor and the train speed measuring system of the invention adopted, the reliability and accuracy of speed measurement of the train can be improved.

The invention discloses a vehicle speed matching method based on linear array CCD image, including the following steps that: two linear array CCD video cameras are respectively arranged on two sections positioned on a road at intervals; both linear array CCD video cameras face downwards and are provided with infrared ray laser source; the frame rate of each CCD video camera is more than 1,000 frames per second, and the resolution is more than 1,024 lines; correlation matching computation of the image sequences obtained by both linear array CCD video cameras is carried out to obtain the time difference of the same object in two image sequences; finally, by means of the time difference and the distance between both cameras, the driving speed of an object vehicle can be obtained according to the velocity formula V=s / t. The invention is suitable for vehicle speed measurement at a traffic station, etc.

The invention discloses a multi-target parameter measurement method suitable for a continuous wave perimeter monitoring radar, belongs to a signal processing technology, and specifically relates to amethod for spectrum correction, multi-target pairing and parameter measurement of a continuous wave radar. A method is provided which solves the problem of distance and speed coupling brought by beatoperation, solves the problem of spectrum correction of ''Range-Doppler'' domain (RD domain) data, so as to realize high-precision distance measurement and speed measurement, and can accurately perform target pairing in a multi-target situation. Aiming at a continuous wave perimeter monitoring radar, an efficiency multi-target distance and speed measurement algorithm with high precision is provided; in the aspect of spectrum zooming, the limitation that traditional spectrum zooming or correction operation is only performed on a one-dimensional distance spectrum is broken through, improvement of speed measurement precision does not just rely on increase of the FFT point number or distance spectrum zooming, but takes full advantage of a result of MTD at the same time, and thus the efficiencyof the algorithm and the utilization rate of information are improved.

The invention relates to a device and method for detecting light-screen parameters of light-screen targets. At present, a method for measuring the distance between two light-screen surfaces is low in measuring precision, the two light screen surfaces are not strictly parallel, high-precision measurement on a shot flying speed cannot be realized, and a light-screen target and a laser target in use need regular detection on the flatness of the screen surfaces and the parallelism and the distance between the two screen surfaces. The detecting device for detecting the light-screen parameters of the light-screen targets is characterized by comprising a datum plane assembly and a distance-measuring datum plane assembly, wherein the distance-measuring datum plane assembly comprises a second laser plane projector, a second plane normal laser projector and a high-precision laser distance measurer, which are arranged on a second precise platform; the datum plane assembly and the distance-measuring datum plane assembly are oppositely arranged in parallel at equal height. The detecting device and the detecting method disclosed by the invention can be used for overcoming the defects that the measurement for the distance between two targets has large errors, that the light-screen surface parallelism cannot be ensured, and the like.

The invention relates to a method and a system for improving the positional accuracy of a Beidou navigation system by using photon pairs with a quantum entanglement feature. The method includes following steps: 1), producing the photon pairs with the quantum entanglement feature; 2), respectively sending the photon pairs with the quantum entanglement feature produced in step 1) to a first clock and a second clock in need of clock synchronization and recording the time from photons to the first clock and the second clock; 3), enabling two photons to pass through a beam splitter, and acquiring the time from the photons to the beam splitter; 4), acquiring a timing difference of the first clock and the second clock; and 5), adjusting the timing difference of one of the clocks according to the acquired timing difference of the two clocks so as synchronizing the two clocks. By the aid of the method and the system, the positional accuracy of the Beidou navigation system is evidently improved.

The invention proposes a high-speed target Doppler velocity measurement method based on the least square method, and the main idea of the method is that the method comprises the steps: determining a radar, supposing that there is a high-speed target in a radar detection range, and calculating the fuzziness of the high-speed target in the radar detection range and the blind speed of the high-speedtarget in the radar detection range; calculating a rough measurement value of the radial speed of the high-speed target in the radar detection range according to the fuzziness of the high-speed targetin the radar detection range and the blind speed of the high-speed target in the radar detection range; calculating a blind speed matrix of the high-speed target in the radar detection range according to the rough measurement value of the radial speed of the high-speed target in the radar detection range, and calculating the Doppler fuzziness root number vector of the high-speed target in the radar detection range based on the least square method; and calculating the real radial speed of the high-speed target in the radar detection range based on the Doppler fuzziness root number vector of the high-speed target in the radar detection range and the blind speed matrix of the high-speed target in the radar detection range.

The invention relates to a high-precision control method for an inertia momentum wheel. The high-precision control method comprises the following steps that: establishing a practical model and a nominal model of an inertia momentum wheel permanent magnetic motor and synchronously measuring signals of a switch type Hall position sensor, thereby calculating and obtaining a rotating speed value of the present sampling period; taking an internal damping moment, a frictional moment and a bearing noise moment, and the like, of the inertia momentum wheel as total disturbance of a system; unifying an output error of the practical model and the nominal model of the inertia momentum wheel permanent magnetic motor to a control input end; and equivalently estimating the disturbance moment of the system by calculating in real time, thereby obtaining an estimating value of the disturbance moment and introducing an equivalent disturbance compensation control voltage at the control input end. According to the method, the disturbance moment of the system of the inertia momentum wheel is effectively restrained, the precision of reproducing a moment instruction of the inertia momentum wheel is obviously increased and the synchronous measuring method is higher in resolution ratio within the full range of rotating speed of the inertia momentum wheel.

The invention relates to a 77GHz millimeter wave corner false-alarm inhibiting system for an automobile active anticollision radar, which belongs to the technical field of radar communication and automobile electron. The system concretely comprises a GPS (Global Positioning System) receiver, an automobile anticollision radar, a geographic information library module, an information comprehensive processing module and a comprehensive display screen, wherein the information processing module is used for carrying out analysis and judgment on data information derived from the GPS receiver, the geographic information library module and the automobile anticollision radar, confirming whether an automobile is close to a corner or is run in the corner or not, obtaining the position information of a potential threatened object and sending an alarm. The 77GHz millimeter wave corner false-alarm inhibiting system for the automobile active anticollision radar has the system working frequency of 77GHz frequency range, small radar size and low false-alarm probability; and an automobile anticollision radar detects or monitors a close object of the turning position and displays the close object on the comprehensive display screen, thereby effectively preventing the automobile from knocking down the object behind the corner.

The invention discloses an automatic train protection system for a monorail, and relates to the technical field of train protection systems. A vehicular controller is respectively connected with a wireless communication terminal, a GPS (global positioning system), a human-computer interaction interface, a transponder transmission module, a speed sensor and a train connector unit, the vehicular controller comprises two sets of secure computer cases and one set of train connector case, the secure computers include an I-series secure computer and an II-series secure computer, and the secure computer cases form a 'square' structure capable of improving system availability. According to characteristics of a monorail train, the train is positioned by the ground in a non-contact manner, three-dimensional space positioning of the train is achieved, and safer and richer signal input and output interfaces greatly improve safety and expandability of the automatic train protection system and adapt to a monorail train system.

The invention provides a method for determining speed measurement error of a Doppler velocity log (DVL) in a strapdown inertial navigation system. The method comprises the following steps: determining an initial position parameter of a carrier by a global positioning system (GPS), acquiring data output by an optical fiber gyro and an accelerator, performing initial alignment on data processing and determining an initial strapdown matrix; acquiring angular motion information and linear motion information of the carrier measured by an inertial component, and performing navigation calculation by a compass method and an inertial navigation method respectively, wherein carrier motion speed information measured by the DVL is introduced into the calculation by the compass method; performing subtraction on two groups of gesture information calculated by the two methods and performing transformation to obtain an azimuth misalignment angle value of the two group of calculated gestures; and performing conversion on the azimuth misalignment angle value to obtain the speed measurement error of the DVL. By the method, the speed measurement error of the DVL can be estimated in the navigation process of the carrier, and the speed measurement precision of the DVL is improved after the result can be compensated to the DVL. The method is simple and easy to operate.

The invention provides a locomotive wheel set idling and sliding detection method and system. The method comprises, after initial alignment is performed, acquiring reference speed, reference acceleration and wheel speed output by a combined inertial navigation system, wherein the wheel speed is determined by speed pulse information collected by a speed sensor; according to the reference speed, the reference acceleration, the wheel speed and a preset formula, computing speed difference, acceleration difference and sliding rate; when the reference acceleration, the speed difference, the acceleration difference and the sliding rate meet preset idling judging conditions, determining that the operating state of a locomotive wheel set is an idling state; when the reference acceleration, the speed difference, the acceleration difference and the sliding rate meet preset sliding judging conditions, determining that the operating state of the locomotive wheel set is a sliding state. Therefore, the locomotive wheel set idling and sliding detection method achieves the effects of detecting whether the locomotive wheel set is idling or sliding to well ensure train safety.

A method of correcting a velocity measurement error of a Doppler velocity log is disclosed. A correcting model based on the method can utilize attitude information provided by an inertial navigation system (INS) to correct a velocity measurement error caused by carrier pitch angle changes for the Doppler velocity log (DVL), thus increasing precision of INS / DVL combined navigation. In the model, an influence of a nonzero pitch angle on the velocity measurement error of the DVL is taken into account, and influences of a carrier pitch angle change in a time difference between DVL signal emitting time and signal receiving time, and lever arm effects caused by a pitch angle velocity and a lever arm at a DVL mounting position on the velocity measurement error of the DVL are also taken into account, and therefore the built error correcting model is more accurate. The carrier attitude information provided by the navigation system is substituted into the velocity measurement error correcting model for the DVL, data acquisition is simple and rapid, and effective error correction can be achieved.

The invention discloses a track traffic rotation speed sensor based on the magnetism principle. The track traffic rotation speed sensor based on the magnetism principle comprises a base, a shaft system assembly and a cover. Circuit modules and a magnetic code disc are arranged in a cavity defined by the base and the cover. The magnetic code disc is fixed to the upper end of a rotary shaft. A plurality of magnetic pole pairs are arranged on the circumference of the magnetic code disc. The N poles and the S poles of the magnetic pole pairs are distributed alternately. Each circuit module comprises a magnetic sensitive element and a circuit board. The magnetic sensitive elements are right opposite to the magnetic poles of the magnetic pole pairs of the magnetic code disc. As the magnetic code disc rotates, the N poles and the S poles of the magnetic pole pairs approach the magnetic sensitive elements alternately, the magnetic sensitive elements sense the magnetic field changes, and signal processing circuits in the circuit boards process sensing signals and output pulse signals. The multiple circuit modules can be fixed in the cavity, each circuit module is internally provided with two magnetic sensitive elements with the phase difference of 90 degrees. The track traffic rotation speed sensor based on the magnetism principle can measure the rotation speed of the rotary shaft accurately, and is good in pollution resistance, not prone to aging and simple in structure.

Provided is a low-interception velocity measurement method which comprises the following steps: 1) generating a binary phase coded signal, the carrier frequency of which is fci, in the ith FHI, and emitting the binary phase coded signal, receiving an echo signal of a target to be measured and carrying out frequency mixing on the echo signal to obtain an intermediate frequency signal; 2) mixing the intermediate frequency signal with the signal itself, and carrying out band-pass filtering to obtain a square rate signal; 3) carrying out sampling on the square rate signal obtained in the step 2) and obtaining a digital sequence signal after sampling; and 4) carrying out Fourier transform on the digital sequence signal, obtaining normalized frequency corresponding to frequency domain peak value according to a frequency spectrum signal obtained after the Fourier transform, and calculating the velocity of the target to be measured. The invention also discloses a low-interception velocity measurement radar device which comprises an antenna, a controller, a two-phase modulator, a frequency synthesizer and a reception branch. The low-interception velocity measurement method and radar device effectively reduce the probability of the velocity measurement radar being intercepted by an electronic dog; and meanwhile, system sampling rate can be effectively reduced, and velocity information of a motion target is kept.

The invention discloses a signal processing method and device of apulse Doppler radar and a storage medium. The method comprises the steps of transmitting a pulse group in a circulating manner, wherein the pulse group comprises a plurality of transmission pulses with different frequency codes; receiving an echo signal returned in response to the pulse group; performing matched filtering on the echo signal in sequence by using a matched filter bank, and performing separation of different range ambiguity regions; performing range walk correction on each range ambiguity region by using a Keystone transformation method; traversing Doppler fuzzy numbers in a preset range to perform Doppler fuzzy number compensation; and according to the peak value of the coherent accumulation result after Doppler fuzzy number compensation, determining a corresponding Doppler fuzzy number, and completing Doppler ambiguity resolution. According to the method, range ambiguity and Doppler ambiguity can be solved at the same time under the condition that repetition frequency is not changed, coherent processing capacity of echo pulses in the whole residence time can be achieved, Doppler resolution is improved, more signal-to-noise ratio gains are brought, and distance measurement precision and speed measurement precision are improved.

The embodiment of the invention discloses a fused speed measurement processing method and device based on accelerometers. The method comprises the following steps: enabling three accelerometers to bearranged in three different angles, and according to the angles of the three accelerometers, calculating an accelerated speed of a current train; measuring a speed of the train at a current moment through a speed sensor; and performing fusion calculation on the accelerated speed and the speed to obtain a target speed of the current train. Through adopting a speed measurement judging mode based onthe accelerometers, the method disclosed by the invention fuses speed measurement processing of the speed sensor and the accelerometers; the interference of an external environment to the precision ofa speed measurement system does not need to be taken into account; the speed measurement precision and usability of signal equipment are improved; and unnecessary loss is avoided.

The invention discloses a normally closed type intelligent catcher belonging to the technical field of mine hoisting equipment protection. The catcher is provided with a car blocking column with a buffer action on a well wall. A performing structure for lifting and descending the car blocking column is arranged on the car blocking column. Two buffers are symmetrically arranged on a track in front of the car blocking column and are connected with the car blocking column. The catcher also comprises a speed sensor for detecting the speed of a mining car, a controller connected with the speed sensor, a monitoring case having the functions of displaying the state of the car blocking column and alarming and connected with the controller, a power supply and a switch which are connected with the controller, wherein the power supply is connected with the performing structure; the speed sensor is arranged on a depth indicator for indicating the lifting / descending depth of a mining car by a winch; an upper limit sensor and a lower limit sensor for detecting the state of the car blocking column are arranged on the performing structure and are connected with the controller. The invention has the advantages of improving the safety in the transportation process of an inclined shaft.

The invention discloses a parallel streamlined processing method for LFMCW array radar signals, and aims to provide a processing method which can improve the operation efficiency, save hardware resources and lower the system cost. The technical scheme is implemented as follows: a digital receiving unit adopts a time-sharing sampling chip A / D which integrates an embedded programmable low-noise amplifier (LNA) and a programmable anti-aliasing filter for receiving radar array signals; digital processing is carried out by adopting a digital interpolation synchronization algorithm, and then a fieldprogrammable gate array (FPGA) is adopted to carry out fine topology on a space-time frequency three-dimensional complex signal processing algorithm, so that the calculation amount is reduced; a space-time frequency three-dimensional joint processing target detection mathematical model is constructed; finally, the processing module algorithms are subjected to decomposition fusion, and a completeparallel streamlined structure of the fine topology is constructed; and the signal processing algorithms are mapped to the FPGA, and parallelization, streamlined and pipeline treatment are carried outon the linear frequency modulation continuous waves ( LFMCW).

A surface topology map technique and apparatus are disclosed for determining calcaneus thickness for imaging quantitative ultrasound measurements; improving measurement accuracy, particularly in in vivo applications.

The invention discloses a Doppler radar-assisted GPS / INS vehicle speed measuring method. According to the method, different geographical coordinate systems and coordinate transformation matrixes are defined; a filtering method such as wavelet transform is adopted to preprocess radar / IMU output data; corresponding error models are established according to the working principles and main error mechanisms of different sensors; and therefore, main ways and methods that affect the accuracy of speed measurement and improve the safe driving performance of vehicles are analyzed. Finally, a neural network-based combined system training and prediction model is provided. According to the method, the advantages and disadvantages of different measurement methods are fully considered; the different error models are established; corresponding speed estimation models are independently selected according to the state changes of the vehicles during actual driving processes; and therefore, the improvement of speed measurement precision can be realized, and a system has better robustness.

A receiver for a multi-mode navigation system includes a base band unit and a calculation unit. The base band unit is configured for allocating resources to positioning satellites in each of one or more navigation system, and tracking the positioning satellites with the allocated resources to obtain satellite information of each of the positioning satellites. The satellite information includes one or more of pseudo-ranges, position coordinates, velocity information, and frequency information of the positioning satellites. The calculation unit is configured for receiving the satellite information from the base band unit, evaluating the positioning satellites in each navigation system, and determining a positioning parameter of the receiver. Determination of the positioning parameter includes calculating a position and a velocity of the receiver based on the satellite information according to a least square algorithm.

The invention discloses a high-speed target detection method based on P / D band radar signal fusion processing. The method comprises the following steps: a transmitted waveform of a P-band working radar is selected as a pseudo-random binary phase shift keying pulse signal and a transmitted waveform of a D-band working radar is selected as a linear frequency modulated continuous wave signal; transmitted waveform parameters of the pseudo-random binary phase shift keying pulse of the P-band radar and the linear frequency modulated continuous wave of the D-band radar are determined; the corresponding working waveforms of the P-band and D-band are transmitted and received cooperatively; echoes of the transmitted signals of the P-band and D-band radars are processed; and the signal processing results of the two waveforms are fused to obtain the distance and velocity information of a target. According to the invention, the problem that velocity detection ambiguity of the high-speed target is likely to occur under the D-band radar is solved; and the accuracy of target ranging and velocity detection is improved.

The invention discloses a UHF waveband multi-channel radar radial speed detection method based on a tensor product. The problems that in the prior art, the low-speed moving target detection probability and radial speed estimation precision are low under the condition that an antenna bearing aperture is limited are mainly solved. According to the implementation process, the method comprises the steps that a tensor product operation is carried out on data vectors received by a radar to obtain data with a virtual aperture expanded; a tensor product operation is respectively carried out on search clutter guide vectors and search target guide vectors to obtain search clutter guide vectors and target guide vectors with the virtual aperture expanded; the search clutter guide vectors and target guide vectors with the virtual aperture expanded are tensioned into search signal subspace; the radial speed of a target is estimated through signal subspace fitting. The method has the advantages of being high in low-speed object detection probability and radial speed estimation precision, and is used for estimating the low target radial speed under the condition that the antenna bearing aperture is limited.

The invention discloses a traffic information acquiring system based on linear CCD imaging, which comprises two linear CCD cameras spaced at a certain interval and disposed on a support above two road sections, each is provided with an auxiliary lighting headlight; and a plurality of auxiliary planar array cameras arranged at the position 15 m away from the front of the two linear CCD cameras and respectively located on the middle of a driveway and on the separating line of the driveway, wherein the auxiliary planar array cameras in the middle of the driveway can shoot vehicles passing the road, and the auxiliary planar array cameras in the driveway separating line can perform panning shot. When the vehicles pass the road section, the linear CCD cameras can acquire linear image data of the vehicles passing the road section; the acquired linear image data is processed and analyzed to obtain traffic parameters such as traffic flow, vehicle velocity, traffic stream density and road occupation rate, and the planar array cameras can shoot vehicle image which is analyzed and processed to obtain license plate information.

The invention discloses an Alignment method of shooting line and object moving direction based on linear array CCD camera velocity measurement; a linear array CCD camera, a front laser point light source and a back laser point light source are arranged above a fast-moving object along the object moving direction, the front laser point light source and the back laser point light source are irradiated on the surface of the object to generate a front light spot and a back light spot; by continuously shooting pictures when the surface of the moving object is located in a certain place of the shooting line and shooting pictures again after the object moves a certain distance, the invention finds out whether the pictures of the surface of the object at a certain place appear in the pictures shot later, if so, the object moving direction is in accordance with the shooting line and stays the same without deviation, otherwise, deviation occurs to the object moving direction, then the linear array CCD camera needs turning a small angle, so that the shooting line is in accordance with the object moving direction, thus maintaining the precision of the camera in measuring the object velocity, and greatly improving the velocity measurement precision of the linear array CCD camera.

The invention relates to the technical field of motor speed measurement by a four-way parallel sampling T-method motor speed measurement method, and specifically relates to a method for obtaining the rotational speed of an incremental photoelectric encoder applied to a servo system at low speed operation. First, the incremental photoelectric encoder Output the square wave signals of phase A and phase B to the signal processing controller, set the trigger signals of the four timers as: rising edge of phase A, falling edge of phase A, rising edge of phase B, falling edge of phase B, and then timing The timer starts counting the high-frequency clock pulse signal with a cycle of T, when the timer encounters the same A and B phase change edges again, it records the number of high-frequency clock pulse signals, and at the same time clears the counter and starts the next count cycle, and then calculate the motor speed according to the number M of high-frequency clock pulse signals, and update the speed every time the timer finishes counting the high-frequency clock pulse signals, thus completing the speed measurement of the motor. The speed measuring method of the invention has high speed measuring precision when measuring the low speed of the motor, low dynamic speed measuring error and high dynamic response speed.

The present invention discloses a hub motor. It includes motor main case and motor cover which are mutually fixedly connected, on the main shaft of its shaft centre the stator left and right end covers fixedly connected with said motor are cover-mounted, said motor can be covered by the left and right end covers of stator and mounted in the interior of motor main case, in the interior of rotor iron core of motor a rotor liner is mounted, and between said motor and planetary reducing mechanism a roller clutch formed from clutch outer ring, roller retaining frame, roller and lefa spring is set, and said planetary reducing mechanism is a double-geared speed reducing one formed from driving gear, planetary gear on planetary spindle, gear III, bearing gear and internal gear, the internal gear is fixedly connected with motor cover so as to implement power transmission.