52results about How to "Improve output" patented technology

A cooling structure of a rotor for a rotary electric machine includes: a rotatable shaft that is configured to supply coolant that flows inside the shaft to outside the shaft; a rotor core that is fitted onto the shaft and fixed thereto, and has a coolant flow path for flowing the coolant, in an axial direction of the rotary electric machine, that is supplied from the shaft, and is formed of a plurality of magnetic plates stacked together in the axial direction of the rotary electric machine; and a coolant impermeable nonmagnetic member that is provided in the rotor core near an inner circumferential surface on a radially outer side in a radial direction of the rotor core, of the inner circumferential surface of the coolant flow path.

Method and arrangement for controlling actual torque in a driveline of a land-based motor vehicle. Taking a direct, non-predictive measure of actual torque induced in the driveline of the motor vehicle and ascertaining a magnitude thereof. Ascertaining a magnitude of a torque demand being requested to be instituted in the driveline. Comparing the ascertained magnitude of the direct, non-predictive measure of actual torque taken is compared with a driveline-configuration dependent maximum torque threshold; in this way, an over-threshold torque condition is ascertained in the driveline based thereupon if such a condition exists. In multi-geared transmissions, the variability can be accounted for through the engagement of different ratio gears. In the event that such an over-threshold torque condition is detected, at least one of several possible over-threshold torque compensating effects will be applied to reduce or diminish the excessive portion of the torque. The remedial measure may be chosen from several possibilities including reconfiguring the variable ratio transmission and / or reducing powerplant-induced driveline torque.

The invention relates to a data generation method based on an automatic encoder network and a generative adversarial network. The data generation method provided by the invention can effectively solvethe problem that an anti-fraud model is difficult to construct due to sample imbalance in electronic transaction. The sample characterization capability of an automatic encoder is used; an input sample code is combined with random noise to serve as generator input; guidance information is added to prior noise, the coded noise can fairly allocate the probability of generation of each sample, the capture capability of the generation model for edge distribution data is effectively improved, and the model provides a new technical support scheme for solving the problem of fraudulent sample deficiency and has certain practical value.

The invention provides a gate drive circuit, which comprises multiple stages of gate drive units, wherein each stage of gate drive unit comprises first to sixth switching elements. A latch consists of the third switching element, the fourth switching element, the fifth switching element and the sixth switching element and is used for first-stage pull-up and second-stage pull-up of voltage of control signals output by the latch by receiving the two upper stages of gate drive signals and one upper stage of gate drive signals so as to increase the voltage output by the latch to the first control terminal of the first switching element. According to the gate driving circuit provided by the invention, in a period of time that the gate drive unit outputs the voltage of the gate drive signals, the latch outputs a very high voltage to the control terminal of the first switching element to improve the output capacity of the first switching element, therefore, under the premise of ensuring that the output voltage of gate drive signals is high enough, the size of the first switching element is further reduced, and the power consumption is reduced.

The present invention discloses intraframe predictive coding method and system relating to video picture compression technique. It contains to each band at least to make twice encoding, each time coded object being different, it capable of being said band original data or said band type conversion result, selecting out best result outputting, at the same time outputting one identify for indicating best code efficiency object type, if it being converted data then needing indicating type, capable of using number to distinguish the conversation type. In decoding, said invention first makes normal decoding to band, then according to identify corresponded conversion type to make relevant inverse conversion to decoding result.

The invention discloses a semiconductor opening switch (SOS)-based ultra-wide spectrum pulse generator with 100 kHz pulse frequency, and belongs to the technical field of pulse power. The pulse generator comprises a charging power supply, a primary low-voltage quick pulse generation unit, a magnetic saturation linear transformer, a pumping capacitor, an SOS and a load in sequence. A plurality of primary low-voltage quick pulse generation units are uniformly arranged at the periphery of each single module and thus the primary current output capacity is increased. A trigger controls the on and off of a metal oxide semiconductor field effect transistor (MOSFET) to generate scores of ns to hundreds of ns pulse with a repeated frequency of 100kHz. The magnetic saturation linear transformer is formed by connecting a plurality of single modules in series. On one hand, the output voltage can be increased by increasing the series number of the single module; on the other hand, after the transformer core is saturated, the secondary side is a coaxial squirrel cage structure, the inductance of a reversed pump is reduced, the time of a reversed current pump can be shortened, an SOS quickly cuts off current, and a high-voltage ultra-wide spectrum pulse is generated in the load. A mismatching charging mode from a primary energy storage large capacitance to a secondary pump small capacitance is used to obtain a high voltage on a pumping capacitor; and on the other hand, under the same target voltage, the series number of the single module is reduced, the inductance of the reversed pump is reduced, and 100kHz ultra-wide spectrum pulse generation is promoted.

A bicycle electric generator is provided with a power generation unit and a controller. The power generation unit includes a rotor arranged to rotate and a stator with a coil arranged to produce a plurality of electrical output states in which a number of turns of the coil that are used differs depending on a rotating state of the rotor. The controller is configured to selectively control the electrical output states of the power generation unit in accordance with the rotating state of the rotor of the power generation unit.

The invention discloses a target detection method and device. The method comprises: acquiring an initial image; inputting the initial image into a pre-trained deep convolutional network model to obtain a second image which is output by the deep convolutional network model and marked with an interest target candidate box, the interest target candidate box possibly having an interest target contained in the initial image; wherein the deep convolutional network model is obtained by training in advance by using a training image marked with real information of an interest target; wherein the inputof the deepest convolution layer in the deep convolution network model comprises an output result of at least one other convolution layer and a convolution result obtained by performing convolution operation on an output result of an adjacent convolution layer of the deepest convolution layer and network parameters. According to the method and the device, the position accuracy of the candidate boxof the interest target contained in the output initial image is improved.

The invention discloses an energy dispatching method of a tramcar hybrid energy storage system. The energy dispatching method comprises the following steps of (1) acquiring train driving characteristic, determining train traction load energy, performing fuzzy control on a train power monitored in real time and a supercapacitor charge state to obtain charging rate of a lithium battery energy storage unit, and calculating maximum charge-discharge power of the two energy storage units; (2) calculating capacity of the hybrid energy storage system, calculating respective pre-allocation energy of the two energy storage units in the hybrid energy storage system by a lowpass filtering algorithm according to a start and stop load energy curve of a train; and (3) performing secondary correction on the pre-allocation energy according to the respective real-time charge states of the two energy storage units, multiplying a secondarily-corrected energy value by an adjustment coefficient by taking maximum charge-discharge energy and the maximum charge-discharge power obtained in the step (1) as a constraint condition, and determining energy allocation of the two energy storage units. The energy dispatching method is used for energy recycling of the tramcar, and the energy utilization efficiency and the system stability are improved.

The invention belongs to the technical field of inertial navigation systems and satellite navigation systems, and particularly relates to a lever arm error calculation method based on a nonlinear tracking differentiator in the SINS / satellite navigation system gravity measurement process. The lever arm error calculation method comprises the steps that specific force of an SINS in a carrier coordinate system is obtained; a velocity vector of a geographic coordinate system relative to an earth coordinate system is obtained; the angular velocity from the geographic coordinate system to an inertial coordinate system is obtained; the angular velocity of a carrier relative to the geographic coordinate system after navigation calculation is conducted on the carrier; the lever arm distance between the SINS and a satellite navigation system in the carrier coordinate system is obtained; an expression in the geographic coordinate system and the relation between the lever arm distance and the angular velocity are obtained; the differential value of a skew symmetric matrix is obtained; a lever arm error value is calculated. According to the lever arm error calculation method based on the nonlinear tracking differentiator, the problem that traditional direct lever arm error processing brings in noise and is poor in accuracy, when the nonlinear tracking differentiator is configured, the nonlinear tracking differentiator can be configured as long as parameters are adjusted properly and a reasonable nonlinear function is constructed, and an output result with quite high accuracy is obtained.

A red light color strengthening method of traffic lights includes the steps that a red light image is continuously collected through imaging equipment, and a rough area containing the red light area is marked in the red light image; a historical red light area is initialized, and the historical red light area is as big as the rough area; a time model is provided, a red light presentation color model is built through a red light presentation color module building step, and the duration time is t color; within each t static and t color time, color strengthening is carried out on the rough area through the red light presentation color model and a red light color strengthening step, and the t color and the t static represents the on-line study time of the red light presentation color model and the time that the red light presentation color model can keep respectively; after each time of on-line study, the red light presentation color model is updated. When a rod for supporting the imaging equipment vibrates or shakes, the actual red light area can still be detected under the influence of natural factors, and therefore the color can be strengthened no matter how the representation color of the red light is.

The invention provides an input method and an input system for providing number selection for ordinal words, belonging to the field of input methods. On the basis of the traditional input method, the input system is additionally provided with a word bank containing the ordinal words and measure words and a determination unit for determining whether sequentially continuous ordinal words and measure words exist among characters input by a user or not according to the word bank. The input method comprises the following steps of: firstly, generating the word bank containing the ordinal words and the measure words; acquiring the characters input by the user by utilizing an input interface; determining whether sequentially continuous ordinal words and measure words exist among the characters input by the user or not according to the word bank; and adding the characters in candidate words corresponding to the characters input by the user according to an acquired determined result under the condition that the sequentially continuous ordinal words and measure words exist among the characters input by the user, thereby generating a candidate number and word list. By setting of the word bank containing the ordinal words and the measure words, the numbers and words input by the user and sequencing codes can be effectively identified, and thus an output result is enhanced when the ordinal words and the measure words are in continuation.

The invention provides a modularized vibration absorber, a vibration absorber module and a machine tool. The modularized vibration absorber comprises at least one vibration absorber unit and at leastone connecting assembly. Each vibration absorber unit comprises a vibration absorber single body and a connecting structure, wherein the connecting structure is in a polyhedral shape and arranged outside the vibration absorber single body in a sleeving mode; and any inner wall surface of the connecting structure is in local contact with the outer surface of the vibration absorber single body, thegap between the vibration absorber single body and the connecting structure is fully reduced, and thus the space between the connecting structure and the vibration absorber single body is more compact. Any surface, far away from a plug, on the polyhedral bodies can serve as connecting surfaces, thus any two vibration absorber single bodies can be fixedly connected together through the connecting surfaces of the corresponding connecting structures, multi-degree-of-freedom combination of the vibration absorbers is achieved, when the multiple vibration absorber single bodies which are overlaid through the connecting surfaces are arranged coaxially, the output force can be multiplied axially, the structure is compact, the vibration absorber single bodies only need to be overlaid in the same direction so that the output force can be adjusted, and conveniences and the rapid effect are achieved.

According to a GOA circuit and a display panel provided by the embodiment of the invention, the size of a second transistor and the size of a third transistor are adjusted, so that the influence of afirst clock signal on the potential of a first node can be effectively inhibited, and the output capability of the GOA circuit is further improved.

The invention provides an auxiliary design method for website interface generation. The auxiliary design method comprises the steps of obtaining a website interface design requirement of a user; Analyzing the website interface design requirement based on a deep neural network algorithm, and converting the website interface design requirement into a website interface design requirement document togenerate a standard requirement document; The standard demand document is sent to a user for confirmation; Obtaining a confirmation instruction of a user; Automatically generating related layouts according to the standard demand document, selecting most matched data or components from a database through an algorithm to perform intelligent layout, and intelligently setting a transition page and a customized page for switching among the layouts; And sending the website interface design scheme to the user. According to the auxiliary design method and system for website interface generation provided by the invention, the modification requirement proposed by a user by using natural voice can be analyzed, automatic modification is realized, and the output result of artificial intelligence is improved by continuously learning human design and automatically improving and improving.

The present invention discloses intraframe predictive coding method and system relating to video picture compression technique. It contains to each band at least to make twice encoding, each time coded object being different, it capable of being said band original data or said band type conversion result, selecting out best result outputting, at the same time outputting one identify for indicating best code efficiency object type, if it being converted data then needing indicating type, capable of using number to distinguish the conversation type. In decoding, said invention first makes normal decoding to band, then according to identify corresponded conversion type to make relevant inverse conversion to decoding result.

Apparatus for moving logs (2) within cutting-off machines. The apparatus includes a platform (1) with one or more longitudinal channels (10) inside which the logs to be cut are disposed, and a pusher device (3) acting in correspondence of the platform (1) and intended for acting on the back of the logs (2) to push them, along respective guide channels (10), toward a cutting device (4) disposed downstream. The apparatus includes a clamping device (5) disposed intermediate between the pusher device (3) and the cutting device (4) and intended for engaging the logs (2) which are pushed along the channels (10) of platform (1) by the pusher device (3) and driving them along up to the cutting device (4).

The invention discloses a Depthwise fast convolution system based on image processing. The Depthwise fast convolution system comprises an image pixel reading unit and a DW convolution calculation unit. The image pixel reading unit can store an image to be processed into the storage unit in a multi-dimensional form; and the DW convolution calculation unit reads the image data in rows and then performs multiply-accumulate calculation with the convolution kernels of the three rows to obtain calculation results of all pixel points. According to the system, a rapid algorithm of image convolution isrealized, the process of repeatedly reading the image storage unit can be reduced, the calculation space used by convolution is reduced, and the result of image convolution can be rapidly and accurately obtained.

The invention discloses a clamp with a linear motor. The clamp comprises a machine station, wherein a linear motor is arranged on the machine station, and comprises a primary stage and a secondary stage; the clamp is characterized by further comprising at least one toggle mechanism; the toggle mechanism comprises a first knuckle arm and a second knuckle arm which are pivotally connected with each other; connecting bars are pivotally arranged at the at least one end of the secondary stage; the other end of each connecting bar is pivotally connected to a pivotal point where the first knuckle arm and the second knuckle arm are connected; the free end of the first knuckle arm is pivotally fixed on the machine station; at least one sliding chute is formed in the machine station; slide blocks capable of moving along the sliding chute is arranged in the sliding chute; the free end of the second knuckle arm is pivotally connected to the slide blocks. The clamp provided by the invention felicitously utilizes the characteristics that after the toggle force-amplifying mechanism exceeds a critical point, power output can be greatly improved, and reliable self-locking and pressurizing can be realized, has a compact structure and comparatively complete functions, and is obvious in force-amplifying effect and high in efficiency.

The invention relates to the technical field of battery charging, in particular to a control method of a dual battery charging system. The dual battery charging system comprises a charging device, a first battery and a second battery. The control method includes the steps of judging the state information of the charging device, if the first trigger condition is satisfied, controlling the chargingdevice to charge the second battery with high power, otherwise, controlling the charging device to charge the second battery with low power; judging the state information of the first battery; if thesecond trigger condition is satisfied, controlling the first battery to charge the second battery with high power, whereas controlling the first battery to charge the second battery with low power. The invention can realize that both the charging device and the first battery can charge the second battery by charging with high power or low power, so that the energy of the charging device and the first battery is fully utilized, the output capacity of the system is improved, the waste of the charging energy is reduced, the performance-price ratio of the system is improved, and the hardware costis not increased.

The invention relates to the technical field of computer data processing, in particular to a new and old kinetic energy conversion performance evaluation method and a device based on a BP neural network. The invention provides the new and old kinetic energy conversion performance evaluation method based on the BP neural network. The method comprises the steps of constructing an index system and performing centralized and standardized processing on collected data to form a training set and a test set; training the BP neural network model to obtain optimal BP neural network model parameters; determining a common index reference point through variable connection weight calculation and variable sensitivity analysis results; calculating the connection distance between the pre-estimated index and the index reference point through an improved K-means algorithm, obtaining the BP neural network model with the input variable importance through the strength, and obtaining the score and sequence of the new kinetic energy development level of the national key city through the output; and re-judging the new and old kinetic energy conversion performance evaluation according to the performance division basis and the new kinetic energy development level score and ranking.

The invention relates to an airport passenger unit income, cost and profit calculating and analysis method. The airport passenger unit income, cost and profit calculating and analysis method includes the steps of classifying 68 classes of passengers in an airport; calculating apportioning coefficients of the 68 classes of passengers; calculating the unit passenger income; calculating the unit passenger cost; calculating the unit profit of a certain class of passengers; conducting summarizing, analysis and the like. According to the method, calculation is carried out according to the airport aviation passenger actual handling capacity, and results calculated under the standard conditions are related to the income and the cost. As the cost shared by various classes of passengers for bearing different cost driver selections directly influences the unit passage income condition, the income, cost and profit system output result in the passenger layer can be perfected; meanwhile, as the policy test result is increased to be brought into the system output result, related financial accounting can be carried out from the realistic perspective, and studying and judgment can be carried out from the future management perspective.

The invention relates to an assembly for a turbojet engine. The assembly comprises an outer shroud (36) and an inner shroud (32) being segmented and comprising circumferential clearances between the segments thereof; an annular row of stator vanes connecting the inner shroud (32) to the outer shroud; a drive with a reduction ratio, which drive is intended to be coupled to a fan; and a loop (34) for cooling and for lubricating the drive, which the loop is configured to heat up at least the outer shroud (36) during the operation of the turbine engine such as to circumferentially reduce the circumferential clearances between the segments. The invention further brings forward a method used for controlling the circumferential clearances between segments of the inner shroud of the turbojet engine.

The invention discloses a train power supply cabinet of an electric locomotive. The train power supply cabinet comprises a first switching switch and at least three power supply subcircuits identical in structure. The first switching switch is connected between an alternating-current input power supply and the at least three power supply subcircuits and used for switching out the required power supply subcircuit serving as a connection subcircuit and connected with the alternating-current input power supply, and normal power supply of the electric locomotive is guaranteed. The train power supply cabinet of the electric locomotive improves output capacity, when one or more power supply subcircuits work abnormally, the power supply cabinet can keep the same output capacity, and accordingly normal power supply of all electric loads is guaranteed.

The invention discloses a double rectifier bridge and double stator automobile generator. The double rectifier bridge and double stator automobile generator comprises rotatable rotors, one annular stator core, two group of stator windings and two rectifier bridges, wherein the rotatable rotors can be used for providing a rotating field; each group of stator winding is twined around part of the stator core to form a stator, and magnetic circuits corresponding to the two stators are not overlapped mutually, and each group of stator winding can generate AC voltage according to the rotating field; each rectifier bridge is connected with the output end of the group of stator winding to convert the induced AC voltage in the stator winding into DC voltage; and the two rectifier bridges are connected in parallel so that the DC voltage of the two rectifier bridges are output in parallel. The double rectifier bridge and double stator automobile generator can reduce the rectified voltage ripple so as to improve the output capability of the generator and reduce the interference of the generator to electric equipment of a finished vehicle.

The invention discloses a multi-parameter uncertainty analysis method based on a value set visualization technology. The multi-parameter uncertainty analysis method comprises the following steps: 1, constructing a value set function in a complex form based on a known analytical expression; 2, quickly drawing a corresponding value set graph and a parameter action arrow by utilizing a value set function; 3, selecting key parameters, and drawing a discrete value set and a parameter action arrow; and 4, completing multi-parameter uncertainty problem analysis according to the coverage range and themoving trend of the value set and the action arrows of the parameters. When solving the problem of multiple uncertain parameters, the multi-parameter uncertainty analysis method based on a value setvisualization technology, on the basis of a value set mapping mathematical theory and in combination with a value set rapid drawing skill, provides a visual expression image to clearly present the effect of each parameter on an output result, and corresponding analysis results can rapidly and visually reveal different influences of multiple parameters on the output result, and then optimal parameter configuration is found, and the output result is improved.

The method comprise: when making image / text separation, firstly getting the pixel distribution information about a part of image; then generating the lookup table required by executing the program to erase the background of image; synchronously making the background erasing program for overall image.

The invention discloses a data monitoring method and device, electronic equipment and a computer readable storage medium. The method comprises the following steps: acquiring a data original message; extracting key data in the data original message, and storing the key data in a data set according to a predetermined hierarchy; receiving an abnormal data query operation, and determining abnormal data in the data set; and monitoring abnormal data at regular time. The technical problem that the abnormal data is difficult to comprehensively monitor when the abnormal data is monitored in the prior art is solved.