Intelligent alignment method and assembly for wireless charging of electric forklift and charging system

Abstract

The invention relates to an intelligent alignment method, assembly and system for wireless charging of an electric forklift. The method comprises the following steps: controlling a plurality of image acquisition devices on the electric forklift to scan a plurality of positioning identifiers, and performing coarse positioning based on a scanning result; in response to the completion of the coarse positioning, determining the current attitude information of the electric forklift and entering a fine positioning stage; calculating position coordinates of the image acquisition devices in the world coordinate system based on observation results of the image acquisition devices on the corresponding positioning identifiers; according to the attitude information and a preset position relationship between each image acquisition device and the center of the forklift, converting the position coordinates into center coordinates of the forklift; performing data fusion on the calculated center coordinates of the forklifts based on the position confidence coefficients corresponding to the positioning identifiers to obtain fused position coordinates; and controlling the electric forklift to move based on the fused position coordinates until a fine positioning completion condition is met. By adopting the method, high precision and high robustness of wireless charging alignment of the industrial forklift can be realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of wireless charging, in particular to a smart alignment method for wireless charging of an electric forklift, a smart alignment assembly and a wireless charging system for the electric forklift. BACKGROUND

[0002] As the core equipment of modern logistics and manufacturing industry, the charging efficiency of an electric forklift directly affects the continuity of operation. Wireless charging technology has become an ideal solution to replace the traditional plug-in charging of an electric forklift due to its advantages of safety and convenience. In a wireless charging system, the alignment accuracy between a transmitting coil and a receiving coil is the key to determining the energy transmission efficiency and system reliability.

[0003] At present, the alignment method between the transmitting coil and the receiving coil generally adopts manual guidance, mechanical installation part guidance, single sensor positioning or basic visual mark recognition. However, these methods generally have narrow fault tolerance range, are easily disturbed by industrial site dust shielding and light changes, or have insufficient stability under complex working conditions, thus being difficult to meet the requirements of industrial forklifts on wireless charging alignment accuracy and robustness. SUMMARY

[0004] Therefore, it is necessary to provide a smart alignment method for wireless charging of an electric forklift, a smart alignment assembly and a wireless charging system for the electric forklift to solve the above technical problems.

[0005] In a first aspect, the present application provides a wireless charging method for an electric forklift, which comprises: controlling a plurality of image acquisition devices on the electric forklift to scan a plurality of positioning marks arranged on the ground of a charging area, and performing coarse positioning based on the scanning results; in response to determining that a coarse positioning completion condition is met, determining attitude information of the electric forklift according to the current positioning mark scanning results, and entering a fine positioning stage; In the fine positioning stage, based on the observation results of each image acquisition device on the corresponding positioning mark, the position coordinates of each image acquisition device in a world coordinate system are calculated; according to the attitude information and the preset position relationship between each image acquisition device and the center of the forklift, the position coordinates of each image acquisition device are converted into corresponding forklift center coordinates in a forklift coordinate system; based on the position confidence of each positioning mark, the calculated forklift center coordinates are data fused to obtain a fused position coordinate of the electric forklift; Based on the fused position coordinate, the electric forklift is controlled to move until the fine positioning completion condition is met.

[0006] In one of the embodiments, the coarse positioning completion condition is: The specific image acquisition device successfully scans the specific positioning mark, and an error between a position of the electric forklift calculated based on the positioning mark and a target point in a world coordinate system is less than a coarse positioning threshold.

[0007] In one of the embodiments, the specific image acquisition device is a rear image acquisition device, and the specific positioning mark is a positioning mark arranged at a position corresponding to a rear part of the forklift in the wireless charging area. In addition, the attitude information of the electric forklift is obtained based on a calculation result of the specific image acquisition device on the specific positioning mark.

[0008] In one of the embodiments, the plurality of image acquisition devices further include a left image acquisition device arranged on a left side of a top protection frame of the electric forklift and a right image acquisition device arranged on a right side of the top protection frame. The left image acquisition device corresponds to at least two positioning marks arranged on a left side of the wireless charging area. The right image acquisition device corresponds to at least two positioning marks arranged on a right side of the wireless charging area.

[0009] In one of the embodiments, when the number of the corresponding positioning marks of one of the image acquisition devices is plural, based on observation results of the corresponding positioning marks by the image acquisition device, the position coordinates of the image acquisition device in the world coordinate system are calculated, including: Based on the observation results of the positioning marks by the image acquisition device, decoding information of the positioning marks is determined. Based on the decoding information of the positioning marks, candidate position coordinates of the image acquisition device are respectively calculated. Based on the candidate position coordinates and the position confidence of the corresponding positioning marks, the position coordinates of the image acquisition device in the world coordinate system are determined.

[0010] In one of the embodiments, the step of determining the position coordinates of the image acquisition device in the world coordinate system based on the candidate position coordinates and the position confidence of the corresponding positioning marks includes: A difference degree between the candidate position coordinates is calculated. If the difference degree is not greater than a consistency threshold, the position coordinates of the image acquisition device in the world coordinate system are calculated by weighted average based on the position confidence of the candidate position coordinates. If the difference degree is greater than the consistency threshold, the candidate position coordinate with the highest position confidence is determined as the position coordinates of the image acquisition device in the world coordinate system.

[0011] In one of the embodiments, the step of obtaining the fused position coordinates of the electric forklift based on the position confidence of each positioning mark and the calculated forklift center coordinates comprises: calculating the fusion weight of each forklift center coordinate based on the position confidence of the positioning mark on which each forklift center coordinate is based; calculating the fused position coordinates of the electric forklift by weighted averaging of each forklift center coordinate based on the fusion weight.

[0012] In one of the embodiments, the step of calculating the fusion weight of each forklift center coordinate comprises: calculating the sum of the position confidence of the effective positioning mark corresponding to the forklift center coordinate; the effective positioning mark is the positioning mark that is determined to meet the effectiveness condition when calculating the position coordinates of the image acquisition device; calculating the ratio of the sum of the position confidence and the total number of the effective positioning marks to obtain the fusion weight of the forklift center coordinate.

[0013] In one of the embodiments, the effectiveness condition of the positioning mark comprises: the decoded information is successfully decoded and passes the verification, and the decoded environmental feature information matches the real-time detected environmental data.

[0014] In one of the embodiments, the positioning mark is a stereoscopic two-dimensional code, and the encoded information of the positioning mark comprises a basic positioning layer, a direction information layer and an extended data layer; the basic positioning layer encodes the absolute coordinates of the positioning mark in the world coordinate system, the floor identifier and the check code; the direction information layer encodes the installation orientation information of the positioning mark; the extended data layer encodes the position confidence of the positioning mark and the illumination level of the installation position.

[0015] In one of the embodiments, after the step of obtaining the fused position coordinates of the electric forklift, the method further comprises: calculating the dispersion degree between each forklift center coordinate; if the dispersion degree is greater than a variance threshold, a sliding window filtering algorithm is enabled to smooth and correct the fused position coordinates.

[0016] In one of the embodiments, the precise positioning completion condition comprises: the distance error between the fused position coordinates and the target charging position is not greater than a precise positioning threshold, and the state lasts for a preset number of fused position coordinate calculation periods.

[0017] In a second aspect, the application also provides an intelligent alignment component for wireless charging of an electric forklift, comprising: a spatial pose perception module comprising a plurality of positioning markers arranged on the ground of a charging area and a plurality of image acquisition devices arranged on the electric forklift; a decision control module in communication connection with the spatial pose perception module and configured to execute a wireless charging method for an electric forklift.

[0018] In a third aspect, the embodiments of the application provide a wireless charging system for an electric forklift, comprising a wireless charging component and an intelligent alignment component for wireless charging of an electric forklift, wherein the wireless charging component is in communication connection with the intelligent alignment component.

[0019] One of the above technical solutions has the following advantages or beneficial effects: by deploying a plurality of image acquisition devices and using a plurality of positioning markers, the weakness of a single data source being easily affected by dust or obstruction is overcome, and the anti-interference capability of the system is improved. By designing a two-stage process of coarse positioning and fine positioning, the problem of narrow initial fault tolerance range is solved by coarse positioning guiding the forklift to a large range working area, and in the fine positioning stage, the positioning data from multiple sources, which may exist conflicts or noise, are comprehensively processed by a data fusion algorithm, so that stable and reliable fusion position coordinates are output under complex working conditions, and the alignment stability is fundamentally improved. In summary, the scheme realizes high precision and high robustness of alignment through the cooperation of multi-sensor redundancy, two-stage positioning and intelligent data fusion, and meets the alignment requirements of industrial forklift wireless charging. BRIEF DESCRIPTION OF DRAWINGS

[0020] Figure 1 FIG. 1 is a structural block diagram of a wireless charging system for an electric forklift in an embodiment; Figure 2 FIG. 2 is a schematic diagram of an array charging coil, a plurality of positioning markers and a world coordinate system in an embodiment; Figure 3 FIG. 3 is an architectural schematic diagram of positioning marker encoding information in an embodiment; Figure 4 FIG. 4 is a flowchart of an intelligent alignment method in an embodiment; Figure 5 FIG. 5 is a flowchart of coarse positioning of an electric forklift in an embodiment; Figure 6 FIG. 6 is a flowchart of fine positioning of an electric forklift in an embodiment; Figure 7 FIG. 7 is a structural schematic diagram of a safety protection module in a ground charging unit in an embodiment; Figure 8 FIG. 8 is an installation position schematic diagram of a foreign matter detection unit and a living body detection unit in an embodiment; Figure 9Fig. 1 is a schematic diagram of a hardware structure of a metal foreign object detection system in an embodiment; Figure 10 Fig. 2 is a schematic diagram of a flow of a metal foreign object enhanced detection method in an embodiment; Figure 11 Fig. 3 is a schematic diagram of a flow of wireless charging of an electric forklift in an embodiment. DETAILED DESCRIPTION

[0021] In order to make the purposes, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application.

[0022] Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the present application. The appearance of the phrase in various places in the specification does not necessarily all refer to the same embodiment, nor does it necessarily exclude other embodiments. It is explicitly and implicitly understood that the embodiments described herein can be combined with other embodiments.

[0023] The present application aims to provide an electric forklift wireless charging system capable of high fault tolerance and high-precision automatic alignment, including a wireless charging assembly for wireless charging of an electric forklift, the wireless charging assembly being configured with a comprehensive, multi-level fusion perception safety protection system, and being capable of realizing intelligent linkage of the charging system with an upper management system and adaptive optimization of an energy transmission process. Understandably, the electric forklift wireless charging system is a safe, efficient and intelligent electric forklift energy supply ecosystem.

[0024] In the embodiments of the present application, the wireless charging assembly described above further includes a ground charging unit and a vehicle-mounted receiving unit, as shown in Figure 1 The ground charging unit 7 can include an array charging coil 1, a high-frequency inverter 2, a resonance compensation network 3, a power factor correction module 4 and a control and communication module 5, and the vehicle-mounted receiving unit 8 can include a receiving coil 9, a high-frequency rectifier 10, a DC-DC converter 11, a battery management system 12, a control and communication module 13 and a user indicator 15. Each functional module will be introduced one by one below.

[0025] In some embodiments, the array charging coil 1 can adopt an embedded or surface-mounted design, and its housing can be made of high-strength stainless steel, thereby having excellent pressure resistance, wear resistance, and corrosion resistance. Specifically, the array charging coil 1 can be of DD or DDQ type structure, which helps to generate a more concentrated unidirectional magnetic field, thereby improving coupling efficiency and alignment fault tolerance. At the same time, the coil can be filled with epoxy resin or other insulating materials for sealing and fixing, so as to achieve an IP67 or higher protection level. In terms of conductor selection, the winding turn group can use Litz wire as the conductive material, which can effectively suppress the additional loss caused by skin effect and proximity effect in high-frequency working environment. In addition, the magnetic core part can use ferrite material with high magnetic permeability and low loss characteristics, such as PC95 or N87 series, which can guide the direction of magnetic lines and further optimize the coupling coefficient. It is worth noting that multiple smaller transmitting coils can be integrated inside the ground charging unit 7 to form a coil array system. By selectively activating the transmitting units closest to the vehicle-mounted receiving coil, the system can maintain high-efficiency energy transmission even if there is a certain deviation in the forklift parking position.

[0026] In some embodiments, the high-frequency inverter 2 can adopt a series-parallel resonant topology, which is an SP-type resonant inverter. Considering that the wireless charging of electric forklifts usually needs to transmit a large power of 11 kW or more and a working frequency of 85 kHz, the high-frequency inverter 2 preferably adopts a Class D or Class E high-frequency soft-switching topology based on S i C (silicon carbide) or G a N (gallium nitride) power devices. This combination can fully utilize the advantages of wide-bandgap semiconductor switches, such as fast switching speed and low loss, and help to achieve a system end-to-end efficiency of more than 90%. In comparison, the Class D scheme based on S i C / G a N is more balanced in terms of cost and performance, while the Class E architecture can achieve higher theoretical efficiency.

[0027] In some embodiments, in terms of the resonant compensation network 3, the ground charging unit 7 can be selected in the form of an SS topology, that is, a series capacitor is used as compensation on both the transmitting end and the receiving coil side. The main advantage of this structure is that when the system works at the resonant frequency, even if the load changes, the receiving end can still maintain a relatively stable output voltage. In other words, during the charging process of the forklift, as the battery voltage, internal resistance and other states change constantly, the constant voltage characteristics of the SS topology can reduce the dependence on external voltage stabilizing circuit. In addition, in the frequency band of 79-90 kHz commonly used for wireless charging of forklifts, this structure can also effectively reduce the reactive power in the coil loop, promote more energy to be converted into active power, and thus improve the overall transmission efficiency. On the other hand, since the forklift inevitably has a position offset when actually parked, the mutual inductance between the coils will change at this time, and the output power of the SS topology changes gently under the condition of mutual inductance fluctuation, so it helps to maintain the stability of the charging process and avoid the sharp decline in efficiency due to alignment deviation.

[0028] In some embodiments, in order to meet the demand for wireless transmission power of 11 kW or more, the power factor correction module 4 can adopt a four-way interleaved PFC structure, that is, a circuit design for high-power output is realized through four parallel Boost converters, and the system efficiency is improved and the component stress is reduced through interleaved control strategy. The main advantage of this structure is that it can disperse the total current to multiple parallel units, thereby reducing the current stress of a single power device and improving its heating and loss conditions. At the same time, the phase interleaving of multiple currents can make the ripples cancel each other out, thereby improving the input current waveform quality.

[0029] In some embodiments, optionally, the control and communication module 5 takes the high-performance master control chip STM32H7 series with ARM architecture as the core to process various real-time control tasks. In terms of communication interface, in order to adapt to complex working conditions in industrial sites, the control and communication module 5 can integrate an Ethernet interface to support industrial protocols such as Modbus TCP / IP and OPC UA, so as to realize data interaction with the upper management system. The upper management system can be a warehouse management system WMS, a freight management system FMS, etc. At the same time, the module 5 can also be equipped with Wi-Fi and Bluetooth wireless communication units for flexible wireless connection between the ground equipment and the vehicle-mounted unit.

[0030] The above describes each module unit in the ground charging unit 7 in the embodiments, and the following will describe each module unit in the vehicle-mounted receiving unit 8 in some embodiments.

[0031] As can be seen from the foregoing, the vehicle-mounted receiving unit 8 can include a receiving coil 9, a high-frequency rectifier 10, a DC-DC converter 11, a battery management system 12, a control and communication module 13, and a user indicator 15. Among them, the receiving coil 9 can use a Litz wire as a winding material to suppress high-frequency eddy current loss. The user indicator 15 can include a display device, which as a human-computer interaction interface, can be used to present the forklift wireless charging positioning process, charging state information, etc. in real time, which is not limited here.

[0032] In the energy processing link, the high-frequency rectifier 10 can use a full-bridge topology based on silicon carbide to improve rectification efficiency and reduce high-frequency loss. The DC-DC converter 11 can use a bidirectional Buck-Boost architecture, which is responsible for converting the received DC power into voltage and current parameters suitable for battery charging. It can support both constant voltage and constant current charging modes; the converter 11 is preferably cooperated with the battery management system 12 to start the discharging function only when receiving the instruction of the system allowing discharging.

[0033] The battery management system 12 can be configured to interact with the vehicle system through the CAN bus and other industrial standard protocols to continuously monitor the key parameters of the battery such as voltage, current, temperature, state of charge and health status, and receive instructions from the control and communication module 13 to realize power-on and power-off management and abnormal processing. The control and communication module 13 can carry a high-performance microprocessor, which can manage the charging process and fault reporting, and also can establish a wireless connection with the ground charging unit 7 through Wi-Fi, Bluetooth module, etc.

[0034] It is worth noting that the control and communication module 5 of the ground charging unit 7 can also be equipped with an Ethernet interface to keep data communication with the warehouse management system or the vehicle fleet management system during the charging process. In this way, the system can report the charging progress including the remaining power and the estimated completion time in real time, so that the warehouse management system can intelligently allocate charging tasks according to the vehicle power state. At the same time, with the wireless communication capability of the vehicle-mounted receiving unit 8, the driver can view the allocated charging area through the display screen and automatically trigger the positioning process after driving into the corresponding two-dimensional code area.

[0035] It should be noted that the electric forklift wireless charging system provided by the embodiments of the present application not only includes the wireless charging assembly as described above, but also includes a multi-modal high-precision intelligent alignment assembly, a full-range dynamic safety protection assembly, and a self-adaptive efficient energy transmission assembly. The relationship of these assemblies can be understood as follows: the wireless charging assembly is used to realize wireless charging of the electric forklift, the intelligent alignment assembly is used to ensure spatial position calibration between the ground charging unit 7 and the vehicle-mounted receiving unit 8, the dynamic safety protection assembly is used to perform high-quality safety monitoring during alignment and wireless charging, and the energy transmission assembly can be understood as a hardware topology structure capable of realizing efficient energy transmission after alignment. The electric forklift wireless charging system realizes full-process automation from alignment, charging, safety monitoring to energy transmission through cooperation of multiple assemblies.

[0036] The multi-modal high-precision intelligent alignment assembly, the full-range dynamic safety protection assembly, and the self-adaptive efficient energy transmission assembly will be introduced and described below.

[0037] 1. Multi-modal high-precision intelligent alignment assembly

[0038] The intelligent alignment assembly is a prerequisite for efficient energy transmission. It does not rely on a single sensing source, but through multi-source information fusion, it realizes stable and reliable alignment and positioning of the ground charging unit 7 and the vehicle-mounted receiving unit 8 in a complex industrial environment.

[0039] First, in terms of hardware configuration of the intelligent alignment assembly, it can include a spatial pose perception module and a decision control module.

[0040] The decision control module can be integrated into the control and communication module on the vehicle or the ground, used to run positioning algorithms and calculate relative poses. Of course, the decision control module can also be independent of the control and communication module on the vehicle or the ground, configured as a separate module, which is not specifically limited here.

[0041] The spatial pose perception module can include a plurality of positioning marks arranged on the ground of the charging area and a plurality of image acquisition devices arranged on the vehicle-mounted unit.

[0042] The multiple image acquisition devices are located, but not limited to, the overhead guard and the bottom of the vehicle body. Specifically, the image acquisition devices can be configured as camera assemblies. Furthermore, the image acquisition device can be equipped with at least three cameras, which can be located on the left, right, and rear sides of the overhead guard of the electric forklift, referred to below as the left camera, right camera, and rear camera; of course, a bottom camera can also be installed on the bottom of the vehicle body for other auxiliary positioning purposes, and there are no restrictions here. The intrinsic parameter matrix K, distortion coefficient D, and extrinsic parameters relative to the forklift coordinate system (with the forklift center as the origin), such as the translation matrix t and rotation matrix R, of at least the left camera, right camera, and rear camera have all been pre-calibrated.

[0043] Multiple positioning markers, such as QR codes and barcodes, are used. In some specific embodiments, high-precision QR codes are selected as the positioning markers. When the electric forklift enters the charging area, the decision control module integrated in the control and communication module 13 is activated, and the image acquisition device is used to scan the high-precision QR codes placed on the ground to assist in positioning. Each QR code can be configured to be made with a square size of about 10 centimeters, but it can also be set to a larger or smaller size according to the actual situation. The QR code can be configured to be a high-precision QR code with high contrast characteristics. The number of high-precision QR codes can be set to one or more. In one embodiment, the number of high-precision QR codes is five, corresponding to the four cameras mentioned above. The five high-precision QR codes are respectively installed in the center and surrounding area of ​​the array charging coil 1, such as... Figure 2 As shown, the four cameras, numbered 1-5, are designed to recognize five high-precision QR code patterns and calculate their pose parameters. Specifically, the four cameras cover three zones: the right camera recognizes QR codes 1 and 2, providing information for the upper path; the left camera recognizes QR codes 3 and 4, providing information for the lower path; and the rear camera recognizes QR code 5 for precise docking. By arranging the five QR codes in three zones, the left and right cameras can perform pre-scanning of the driving path, and the rear camera can achieve precise docking, expanding the effective working range compared to a traditional single QR code layout. Of course, in other embodiments, the number of cameras and QR codes, their correspondence, etc., can be configured according to actual needs and are not limited to the above embodiment.

[0044] In some preferred embodiments, in order to ensure the reliability of the image acquisition device under complex industrial lighting conditions, the system also integrates an adaptive multi-spectrum light supplement system. The light supplement system can avoid the influence of uneven lighting, backlight, low illumination and sudden changes in light, so that the camera can continuously and stably capture clear two-dimensional code images in any lighting environment. In this embodiment, each camera, at least the left camera, the right camera and the rear camera described above, can integrate a dedicated light supplement module. The light source can use a combination of infrared LED (such as 850nm / 940nm) and visible light white LED. The infrared LED can be used as the main light supplement source and is invisible to the human eye, which can avoid causing glare to the driver and can penetrate a certain degree of dust and mist. The white light LED is used as an auxiliary light source and is turned on for a short time in a specific low-frequency mode when the system detects extremely low illumination and the infrared light supplement effect is poor, providing additional illumination. Specifically, the system can also include a high dynamic range ambient light sensor arranged at the position of the roof rack to be able to monitor the ambient light intensity and color temperature in real time. The ambient light sensor and each light supplement module can be in communication connection with the decision control module to work cooperatively with the two-dimensional code recognition algorithm.

[0045] The specific light supplement strategy can be: for regular or bright environment: when the ambient light is sufficient, the system automatically turns off the light supplement to save energy, at this time the camera can directly use natural light or ambient lighting for acquisition. For low light or night environment: the ambient light sensor can automatically turn on the infrared light supplement when detecting that the illumination is lower than the threshold. For backlight or strong shadow environment: the system can analyze the image histogram to determine that the scene is high contrast, such as when a forklift enters a dim garage from a bright area, at this time the high-power infrared light supplement can be started to suppress the background light and illuminate the two-dimensional code area. For extreme low light or recognition failure: if the two-dimensional code cannot be successfully decoded for several frames, the system can determine that it is an extreme condition, at this time, the infrared LED and visible light white LED composite light supplement mode can be started. The white light LED flashes in a short pulse form with high brightness, which is accurately synchronized with the exposure time of the camera, and provides sufficient illumination for a moment to capture key details. After shooting, it can be turned off immediately to avoid long-term glare.

[0046] It should be pointed out that in some preferred embodiments, each positioning mark is a three-dimensional two-dimensional code, and multiple identification information can be integrated in the three-dimensional two-dimensional code, one of the multiple identification information can be the illumination level of the two-dimensional code installation position, and in this embodiment, as a preferred solution, the illumination level contained in each two-dimensional code information can be used as a pre-judgment parameter of the light compensation strategy. For example, when the vehicle enters a two-dimensional code area marked as LW5 (which can represent the worst illumination), the system can actively start the strong power infrared light compensation in advance, realizing the upgrade from passive response to active prediction, and further shortening the identification time. This scheme can solve the identification problem under various illumination conditions from daytime to night, from indoor to outdoor half room, etc. through the combination of infrared and white light; taking invisible infrared light as the main light, the light pollution to personnel in the industrial field can be avoided; at the same time, the closed-loop control based on ambient light sensing and image quality feedback enables the light compensation system to dynamically adapt to complex and changeable illumination environments; in addition, the hierarchical power control can also avoid energy waste and meet the energy saving requirements of industrial equipment.

[0047] In this preferred embodiment, as shown in Figure 3 each positioning mark is a three-dimensional two-dimensional code, and the multiple identification information integrated in the three-dimensional two-dimensional code includes a basic positioning layer, a direction information layer, and an extended data layer.

[0048] The basic positioning layer can encode the absolute coordinates of the mark in the world coordinate system, floor information, and a check code. In one embodiment, the world coordinate system can be as shown in Figure 2 , at this time, the array charging coil 1 adopts a 5*5 arrangement, wherein the center of the charging coil in the 3rd row and the 3rd column is defined as the origin of the world coordinate system, the X-axis and Y-axis directions are as shown in the figure, and the Z-axis is perpendicular to the X-Y plane and positive upward. The absolute coordinates (x, y, z) in the world coordinate system, x and y are the plane coordinate values with the world coordinate system as the origin, and z is the height coordinate, with the Z value of the world coordinate origin of the first layer as 0. When the forklift wireless charging needs to be performed in different work areas and different floor heights, the Z value of the corresponding two-dimensional code is matched accordingly. The floor identifier can be understood as an identifier for explicitly encoding floor information, such as B1-A representing the first underground floor in area A and F2-B representing the third floor above ground in area B. The identifier can be mutually checked with the Z coordinate in the absolute coordinates, and this scheme gives the electric forklift a technical means to directly charge at a specified location in the modern WMS system operation scheduling. The check code can be understood as a short check value obtained after CRC16 calculation on the absolute coordinates and floor information, and is encoded together; after scanning and decoding, the system will recalculate the check code and compare it with the check code in the two-dimensional code. If they are inconsistent, the scanning structure is discarded, and the forklift waits until the check is passed.

[0049] Direction information layer, used to encode the installation orientation of the mark, such as Yaw: 180°, representing the orientation of the two-dimensional code is 180°, due south. The direction information layer can be used to assist positioning. When the camera scans a two-dimensional code, if the orientation of the two-dimensional code is known, such as the two-dimensional code position information is to the left or to the right, the relative position between the camera and the two-dimensional code can be calculated more easily. Another role is that, in the embodiments of the present application, the installation orientation of the two-dimensional code should be consistent, such as Figure 2 The orientations of the two-dimensional codes in sequence numbers 1-5 can all be defined as left. When the forklift is driving, if the camera captures the orientation information of a certain two-dimensional code that is not correct, the error can be found intuitively and timely.

[0050] The extended data layer can be understood as an environmental feature fingerprint. The environmental feature fingerprint can encode the environmental information of the point, forming a feature summary of the environment around the installation point of the two-dimensional code. The information in the environmental feature fingerprint can include the illumination level of the two-dimensional code installation position mentioned above, which can be divided into 1-5 levels. The illumination level of 1 is the best, and the illumination level of 5 is the worst, and so on. For example, when the illumination level is LW1, it represents that the installation point has the best illumination; at the same time, it can also include multiple camera observation weight information, still taking Figure 2 Taking the five two-dimensional codes as an example, the five two-dimensional codes constitute a system with different weights. The information of each two-dimensional code includes the position confidence of each two-dimensional code in the system, for example, QR1: 0.75, QR2: 0.9, QR3: 0.8, QR4: 0.92, QR5: 0.95, which represents that the position confidence of the first two-dimensional code QR1 is 0.75. When there is a conflict with the current distance calculated based on QR2, because the position confidence of QR2 is higher than that of QR1, the forklift position calculated based on QR2 is adopted.

[0051] A complete three-dimensional two-dimensional code encoding example is as follows: (x: 0.0001, y: 0.2532, z: 0.3542, LW1, QR1: 0.2, QR2: 0.3, QR3: 0.78, QR4: 0.85, QR5: 0.95, F2-C, Yaw: 180).

[0052] The multi-modal high-precision intelligent alignment assembly proposed in the embodiments of the present application can refer to the intelligent alignment method Figure 4 , which includes: S502, control the plurality of image acquisition devices on the electric forklift to scan the plurality of positioning marks arranged on the ground of the charging area, and perform coarse positioning based on the scanning result. S502, in response to determining that the coarse positioning completion condition is met, determine the attitude information of the electric forklift according to the current positioning mark scanning result, and enter the fine positioning stage. S502, in the fine positioning stage, based on the observation results of each image acquisition device on the corresponding positioning mark, calculate the position coordinates of each image acquisition device in the world coordinate system; according to the attitude information and the preset position relationship between each image acquisition device and the center of the forklift, convert the position coordinates of each image acquisition device into the corresponding forklift center coordinates in the forklift coordinate system; based on the position confidence of each positioning mark, data fusion is performed on the calculated forklift center coordinates to obtain the fused position coordinates of the electric forklift. S502, based on the fused position coordinates, control the electric forklift to move until the fine positioning completion condition is met. By deploying multiple image acquisition devices and using multiple positioning marks, the weakness of a single data source being easily affected by dust or obstruction is overcome, and the anti-interference ability of the system is improved. By designing a two-stage process of coarse positioning and fine positioning, the problem of narrow initial fault tolerance range is solved by guiding the forklift to a large range of working area in the coarse positioning stage, and in the fine positioning stage, the positioning data from multiple sources, which may have conflicts or noise, are comprehensively processed by a data fusion algorithm, so that stable and reliable fused position coordinates are output in complex working conditions, thereby fundamentally improving the alignment stability. In summary, the scheme achieves high precision and high robustness of alignment through the cooperation of multi-sensor redundancy, two-stage positioning and intelligent data fusion, and meets the alignment requirements of industrial forklift wireless charging.

[0053] In some specific embodiments, as shown in Figure 5 and Figure 6 , the whole forklift wireless charging positioning process can be divided into three progressive stages: First, the initial detection stage, when the forklift approaches the charging area, the three cameras of the top guard begin to scan the surrounding environment, and once the two-dimensional code is identified, the display screen can immediately pop up a prompt that the charging area has been entered, at which time the vehicle speed can be appropriately reduced.

[0054] Then enter the coarse positioning stage, the system uses the wide-angle view of the three cameras to estimate the lateral offset and angle deviation between the forklift and the charging coil through the PnP algorithm, when the distance reaches a coarse positioning threshold such as 50 centimeters, the system will prompt that the coarse positioning is completed.

[0055] Finally, the fine positioning stage is dominated by the rear camera, which captures fine markers to achieve the preset position accuracy and angle tolerance, and when the display screen confirms that the fine positioning is completed, the forklift is ready to perform wireless charging.

[0056] In one specific embodiment, as shown in Figure 5As shown, after receiving the charging demand, the forklift can drive to the charging area through manned or unmanned driving system. When the forklift approaches the target charging area, that is, when the camera scans the matched two-dimensional code, the process of the forklift wireless charging positioning enters the coarse positioning stage. As shown in the following Figure 2 In the embodiment, the left and right cameras are installed in the left and right middle parts of the canopy frame, and the rear camera is installed in the rear middle part. Due to the design characteristics of the camera structure and the two-dimensional code position, the two-dimensional codes with serial numbers 1-4 will be read earlier than the two-dimensional code with serial number 5. Therefore, when the left canopy frame camera of the forklift can only read the two-dimensional code with serial number 4, the right canopy frame camera can only read the two-dimensional code with serial number 2, and the rear camera cannot read the two-dimensional code with serial number 5, the first stage of the coarse positioning stage is completed, and the vehicle continues to drive to enter the second stage. When the left canopy frame camera of the forklift can read the two-dimensional codes with serial numbers 3 and 4 at the same time, the right canopy frame camera can read the two-dimensional codes with serial numbers 1 and 2 at the same time, and the rear canopy frame camera can read the two-dimensional code with serial number 5, the forklift can slow down according to the position information but still continue to drive until the rear camera can observe the two-dimensional code with serial number 5, and the distance error between the position of the forklift calculated according to the decoding information of the two-dimensional code and the origin of the world coordinate system is less than a preset coarse positioning threshold. Then, the vehicle stops driving, and the system immediately obtains the attitude rotation matrix of the forklift in the current pose based on the calculation result of the rear camera on the two-dimensional code with serial number 5, which is used for subsequent calculation of the center position of the forklift. After the coarse positioning is completed, the pose information of the left and right canopy frames can be checked, that is, the calculation results of the left and right cameras are compared. The calculation result can be used to confirm whether the left and right cameras have successfully captured the respective two-dimensional codes, and to check whether the poses of the forklift calculated by the left and right cameras are greatly inconsistent, so as to exclude the situation of misidentification or serious obstruction of a single camera. Finally, after the verification is correct, the fine positioning process is entered.

[0057] The purpose of the above coarse positioning is to guide the forklift into the approximate range of the charging area. At this time, although the positioning accuracy is low, the forklift can enter the effective working area of the fine positioning process. Further, as shown in the following Figure 6 The fine positioning can adopt a multi-camera cooperative positioning method and a multi-camera position calculation algorithm, specifically as follows: Firstly, the single camera data is fused, that is, each camera uses the observed multiple two-dimensional codes to obtain the most reliable data for calculating the position of the camera in the world coordinate system, and then independently calculates the pose of the camera in the world coordinate system through the PnP algorithm.

[0058] For the left camera, it needs to observe the 3rd and 4th QR codes at this stage. If only one of the two QR codes is observed, the world coordinate of the left camera is determined based on the decoding information of the QR code. If both the 3rd and 4th QR codes are observed, the world coordinate of the left camera can be determined based on the decoding information of both the 3rd and 4th QR codes (i.e., two candidate position coordinates). If the Euclidean distance of the two world coordinates is not greater than a consistency threshold, it is determined that the displacement of the two world coordinates is consistent. In this case, the two world coordinates are weighted and averaged based on the corresponding position confidence decoded from the two QR codes to obtain the final world coordinate of the left camera. If the displacement of the 3rd and 4th QR codes is inconsistent, the world coordinate of the left camera calculated based on the QR code with higher position confidence is selected as the final world coordinate of the left camera. If neither the 3rd nor the 4th QR code is observed, the corresponding error invalid signal is reported. After determining the final world coordinate of the left camera, the light level of the 3rd and 4th QR codes is read and matched with the actual detected ambient light. If they do not match, the system determines that the identification calculation result for this QR code is unreliable, and the corresponding error invalid signal is reported. If they match, the system determines that the identification calculation result for this QR code is reliable, and the final calculated world coordinate of the left camera is retained.

[0059] The decision-making algorithm for the right camera is consistent with that for the left camera, except that it observes the 1st and 2nd QR codes and calculates the world coordinate of the right camera. The details are not repeated here.

[0060] For the rear camera, since it only observes the 5th QR code, the position confidence of the single QR code is the highest, so single camera data fusion is not needed. The process for calculating the world coordinate of the rear camera can be simplified as follows: decode the 5th QR code to obtain the world coordinate, calculate the relative pose of the rear camera based on the corresponding rotation matrix and translation vector, then calculate the world coordinate of the rear camera, read the position confidence and light level of the 5th QR code, which can be used for subsequent weighted fusion calculation of the center position of the forklift; read the light level and match it with the detected actual ambient light. If they do not match, the system determines that the identification calculation result is unreliable, and the corresponding error invalid signal is reported. If they match, the system determines that the identification calculation result is reliable, and the calculated world coordinate of the rear camera is retained.

[0061] Next, the data obtained from single camera fusion calculation is used to calculate the center position of the forklift, i.e., the coordinates of multiple cameras are unified and converted to obtain the accurate world coordinate of the center point of the forklift. The specific steps are as follows: First, the coordinate systems are unified and converted.

[0062] Each camera has a pre-calibrated camera installation parameter, including a fixed offset of each camera relative to the forklift center, based on the camera installation parameter and the forklift attitude rotation matrix obtained by the foregoing coarse positioning, the world coordinates of each camera calculated in the foregoing are converted to the forklift center coordinates. Taking the left camera as an example, the forklift center coordinates corresponding to the left camera world coordinates are X l , the left camera world coordinates are P l , the forklift attitude rotation matrix is R V , and the fixed offset of the left camera relative to the forklift center is Δ l , then X l =P l -R V ×Δ l . The calculation of the forklift center coordinates based on the right camera and the rear camera world coordinates is similar to that of the left camera, and the specific calculation process is not described here. After calculation, the forklift center coordinates corresponding to the right camera world coordinates X r , and the forklift center coordinates corresponding to the rear camera world coordinates X b are obtained.

[0063] Then, the basic weight calculation and normalization of each camera are performed, and the basic weight of each camera is calculated according to the position confidence of the two-dimensional code used by each camera. Similarly, taking the left camera as an example, if the above single camera data fusion is performed, the number of effective two-dimensional codes currently successfully used by the left camera is 1, for example, only No. 3 two-dimensional code is effective, at this time the basic weight ω l of the left camera is completely determined by this available two-dimensional code, that is, ω l =C3 / 1=C3, C3 is the position confidence of No. 3 two-dimensional code; if the left camera observes two effective two-dimensional codes, and the calculated displacements are inconsistent, then ω l =(C3+C4) / 2, C4 is the position confidence of No. 4 two-dimensional code; if the left camera observes two effective two-dimensional codes, but the calculated displacements are inconsistent, at this time the system will select the two-dimensional code with higher position confidence, such as No. 4 two-dimensional code, then ω l =C4 / 1=C4. It can be seen that the calculation method of the basic weight of each camera can be summarized as ω=S / N, where S can be understood as the sum of the confidence of all two-dimensional codes actually used by the camera to calculate its final world coordinates in this frame, and N can be understood as the number of corresponding two-dimensional codes. Similarly, the calculation method of the basic weight ω r of the right camera is the same as that of the left camera ω l , which is not described here. The calculation method of the basic weight ω bThe position confidence C5 of the 5th two-dimensional code can be understood, and of course, if the number of two-dimensional codes observed by the rear camera is not only one in some other embodiments, the calculation method of the basic weight is consistent with that of the left and right cameras.

[0064] Subsequently, the basic weights ω l , ω r , and ω b are normalized, and the final total weight is obtained. That is, the total weight W = ω l + ω r + ω b , the normalized weight W l of the left camera = ω l / W, the normalized weight W r of the right camera = ω r / W, and the normalized weight W b of the rear camera = ω b / W.

[0065] Finally, the three forklift center position estimation values are weighted and averaged based on the normalized weights, and the final forklift center world coordinate X = W l × X l + W r × X r + W b × X b is obtained, which is the final output of the accurate positioning system. Preferably, C5 is preset as the highest confidence, which can make the rear camera with high confidence play a dominant role in decision-making and ensure the reliability of the accurate parking stage; at the same time, when one camera fails, other cameras can still provide calculation data to build a more reliable accurate positioning system through multi-layer verification and fusion.

[0066] After the final forklift center world coordinate is calculated, the distance error between the current position of the forklift and the target position can be obtained according to the coordinate, wherein the target position can be understood as a preset wireless charging target world coordinate, If the distance error is not greater than a preset fine positioning threshold, for example, not greater than 5 cm, or not greater than the preset fine positioning threshold and lasts for a preset verification period, it can be determined that the forklift fine positioning is successful, and the forklift is controlled to enter the wireless charging process. Of course, if the distance error is greater than the preset fine positioning threshold, the forklift can continue to be guided to move until the condition for entering the wireless charging process is met. Among them, when the forklift continues to be guided to move, for the unmanned forklift, the system can take the current distance error as the deviation amount of three degrees of freedom of X axis (forward and backward), Y axis (left and right) and heading angle (θ), and then generate a shortest path from the current position to the target position. The control system of the forklift can control the speed and angle of the driving wheel and the steering wheel according to the path, so that the vehicle moves accurately along the straight line; or the heading angle can be adjusted first to accurately align the vehicle head with the target point, and then the vehicle is controlled to move straight in the direction of the vehicle body to eliminate the remaining position error. For the manned forklift, the distance error between the current position and the target position of the forklift can be displayed on the display device of the forklift, and then some prompt information is given to make the driver operate the vehicle according to the prompt. At this time, the system can calculate and feedback in real time and continuously until the final prompt of positioning success and please stop charging is displayed, that is, the guidance is completed.

[0067] It should be noted that throughout the fine positioning calculation process, the system can perform a set of abnormality processing and performance optimization mechanism in parallel as a more preferred embodiment. This mechanism can run through various links of data acquisition, calculation and output to ensure the robustness and real-time performance of the system. Specifically: On the one hand, during the data acquisition and single camera fusion stage, the system can perform camera fault detection automatically based on the position confidence of the two-dimensional code. Specifically, the system can continuously monitor the position confidence of each two-dimensional code. If the average position confidence of a two-dimensional code for K consecutive frames (such as K = 4) is lower than a preset fault threshold, for example, 0.5, the system will determine that the two-dimensional code data is unreliable and mark it as a fault, set its weight to zero in the subsequent weight calculation link, and re-normalize the remaining valid weights. For example, if the left camera fails, the normalized weight of the left camera is W l = 0, the normalized weight of the right camera is updated to W r = ω b / (ω r + ω r ), and the normalized weight of the rear camera is updated to W r = ω b / (ω b + ω b ). r b ​); if the number of cameras marked as faulty in the system reaches or exceeds two, the system can automatically switch to a degraded mode in the next calculation cycle, in which it will no longer rely on real-time camera data for fusion, but will make position predictions based on the last 10 frames of historical valid forklift center coordinates through least squares method, and the prediction formula can be: X p =β1X1+β2X2+……+β 10 X 10 , where the coefficients β1 to β 10 are pre-calculated least squares coefficients, the sum of which is 1, and the recent frame coefficients will be configured to be larger to ensure the continuity and trend of the prediction.

[0068] On the other hand, the system can automatically perform data consistency checking, calculating the variance between the forklift center coordinates calculated by the current valid camera (i.e., X l , X r , X b ), if the variance is greater than a preset variance threshold (in one embodiment, the variance is greater than 0.0004m 2 , the variance threshold corresponds to a position standard deviation of about 2 cm), it is determined that the data is significantly inconsistent, in this embodiment, the system can enable a 5-frame sliding window filter to correct the current forklift center coordinates to the arithmetic mean of the last 5 frames of valid center coordinates, i.e., X f =(X 1z +X 2z +……+X 5z ) / 5, to smooth out abnormal jumps and improve the stability of the output.

[0069] In another aspect, preferably, the system is also deployed with corresponding performance optimization strategies to ensure real-time task processing and efficient memory management.

[0070] In terms of real-time performance, the system sets processing priorities for different calculation tasks, with the priorities from high to low being: two-dimensional code detection, decoding, and reading of preset confidence and light level; displacement calculation and coordinate consistency check within a single camera; forklift center position calculation and positioning error evaluation; weight adjustment and fault marking.

[0071] In terms of memory management, the system only caches the effective positioning results of the last 15 frames that meet the caching condition that the average confidence of each camera is not less than 0.5 after each frame of data processing is completed, and adopts a dynamic sliding window mechanism. When the distance error between the current position of the forklift and the target position is greater than 10 cm, a small window of 3 frames can be used to prioritize the calculation speed, and when the distance error is reduced to 10 cm or less, the system switches to a window of 5 frames to prioritize positioning accuracy. At the same time, the system can perform outlier rejection. If the deviation of the forklift center coordinates calculated by a single frame from the historical mean value in the sliding window exceeds 8 cm or other values, the system determines that the frame is an outlier and does not participate in the final fusion calculation and historical cache update. It should be noted that the specific numerical values of the above thresholds can be changed and modified according to actual conditions, and this place is not uniquely limited.

[0072] 2. A safety protection assembly for omnidirectional dynamic.

[0073] The safety protection assembly provides intrinsic safety protection for the wireless charging process. In some embodiments, the omnidirectional dynamic safety protection assembly includes a safety protection module 6 located at the ground charging unit 7. As shown in Figure 7 The safety protection module 6 can include five functional units, namely, a foreign object detection unit, a living body detection unit, a temperature sensing unit, a voltage and current sensing unit, and an electromagnetic shielding unit. The installation positions of the foreign object detection unit and the living body detection unit can be as shown in Figure 8 The safety protection module 6 integrates various sensors to achieve omnidirectional monitoring. Each functional unit in the safety protection module 6 will be introduced one by one below.

[0074] First, the foreign object detection unit, which can be a FOD sensing unit, i.e., an optical fiber displacement sensor. For this foreign object detection unit, the system provides a multi-level foreign object detection scheme, including a basic detection scheme and an enhanced detection scheme based on multi-frequency impedance spectrum analysis. The specific contents of the two schemes are as follows: In the basic detection scheme, the detection unit can integrate multiple independent sensitive coils. It should be noted that these coils are dedicated to environmental perception and do not participate in energy transmission. When metal conductive foreign objects enter the charging area, they will change the magnetic field distribution around the sensitive coils. The system detects this change to identify foreign objects and immediately sends instructions through the control and communication module to take protective measures such as stopping charging or reducing transmission power. At the same time, for the detection of non-metallic foreign objects, the foreign object detection unit is also equipped with a detection unit based on machine vision, which uses an intelligent algorithm to identify features of objects with poor conductivity such as plastic and paper. When such foreign objects enter the monitoring area, the system can accurately identify and start the corresponding protection mechanism.

[0075] In the enhanced detection scheme, preferably, the system deploys a metal foreign object detection system based on multi-frequency impedance spectrum analysis and dynamic learning threshold, the core of which is to create a unique impedance spectrum fingerprint by analyzing the complex impedance changes of the detection coil at multiple frequency points, and then to accurately judge whether a metal foreign object exists and its material type.

[0076] Specifically, as shown in Figure 9 the system hardware includes a composite detection coil array composed of independent excitation coils and induction coils, an independent excitation and sweep signal source generating a specific frequency sweep signal, a high-precision impedance analysis module, a dynamic reference learning and storage module for storing and updating the reference impedance spectrum without foreign objects, and a digital signal processor (hereinafter referred to as DSP) for feature extraction and intelligent discrimination.

[0077] Among them, preferably, the composite detection coil array is tiled in a 5*5 matrix form above the WPT transmitting coil, and specifically, the array can be composed of multiple detection units, each unit adopting a concentric circular close-coupled structure of inner ring excitation coil and outer ring induction coil.

[0078] The above-mentioned specific frequency sweep signal is preferably 10kHz-300kHz, and in a further embodiment, the independent excitation and sweep signal source can be an independent excitation and sweep signal source realized by a direct digital frequency synthesizer chip, which is used to generate a sinusoidal signal swept at 128 frequency points in a logarithmic manner in the range of 10kHz-300kHz, and the excitation is independent of the system main power field.

[0079] The impedance analysis module can use an integrated impedance digital converter to measure and output the complex impedance and phase angle of each frequency point.

[0080] The dynamic reference learning and storage module can be optionally composed of a Flash memory inside the DSP; in a further embodiment, the module can be used to learn and store the full-band reference impedance spectrum when the system confirms that there is no metal foreign object, and can update it smoothly at a rate of 4 hours to compensate for environmental slow changes in a safe state. Of course, the frequency of smooth updating can be configured according to actual needs.

[0081] The digital signal processor DSP, as the core processing unit, can adopt a dual-core architecture, one core responsible for system flow control and the other core dedicated to running the subsequent intelligent discrimination algorithm.

[0082] The working principle of the system is based on the impedance model of metal foreign object eddy current effect. In order to clarify the physical mechanism, a mutual inductance coupling model is abstracted from a single detection unit (i.e. one excitation coil and one induction coil) and the metal foreign object that may exist near it. In this model: the self-resistance and inductance of the induction coil are R s and L s respectively; the metal foreign object can be equivalent to a conductive ring with a radius of r, a resistivity of p, and a thickness of d, and its equivalent resistance R metal and equivalent inductance L metal are determined by its geometric parameters and material properties; the mutual inductance between the induction coil and the metal foreign object is M.

[0083] When an excitation with an angular frequency of w is applied, according to Kirchhoff's voltage law, the following equation can be listed: (Coil inductance loop) (1) (Metal foreign object equivalent loop) (2) The induced eddy current I m in the metal foreign object can be solved from equation (2): (3) Substituting equation (3) into equation (1), the input impedance Z in (ω) of the system can be derived: (4) In order to more clearly observe the impedance change introduced by the foreign object, equation (4) is processed to separate the real part and the imaginary part: (5) From equation (5), it can be seen that the existence of the metal foreign object introduces a frequency-dependent resistance increment ΔR(ω) and an inductance decrement ΔL(ω). Let: (6) (7) Then the input impedance can be simply written as: (8) Equation (8) is the core theoretical basis of the embodiment. It shows that the existence of the foreign object can be perceived by measuring the impedance change of the detection coil. Further, the key innovation of the present application lies in the use of the "multi-frequency impedance spectrum fingerprint identification principle": equations (6) and (7) reveal that the frequency characteristics (i.e. impedance spectrum form) of ΔR(ω) and ΔL(ω) are strongly dependent on the intrinsic parameters R metal and L metal of the metal material.

[0084] For high-conductivity non-magnetic metals (such as copper and aluminum), their Rmetal Smaller, L metal Dominated by geometry, its ΔR(ω) curve usually peaks at a certain characteristic frequency and then decreases, while ΔL(ω) decreases monotonically as the frequency increases.

[0085] For ferromagnetic metals (such as steel), the permeability μ is much greater than 1, resulting in an equivalent L metal Significantly increased, and R metal It may also increase due to hysteresis loss, and the curve shapes of its ΔR(ω) and ΔL(ω) are significantly different from those of non-magnetic metals.

[0086] Therefore, Z is measured by scanning the frequency ω over a wide bandwidth. in (ω) or directly calculate its comparison with the foreign matter-free reference spectrum Z. b The difference ΔZ(ω) = Z(ω) in (ω)-Z b By calculating (ω) = ΔR(ω) + jωΔL(ω), an impedance spectrum reflecting the characteristics of a specific metallic material can be obtained. This spectral line, containing amplitude and phase information, is like a fingerprint of the metal, laying the theoretical foundation for the accurate differentiation of foreign materials.

[0087] Please see Figure 10 The workflow of this enhanced detection scheme may include the following steps: After the system is powered on, it first initializes each module and reads the reference impedance spectrum. If no valid reference is found, it enters learning mode to establish a new reference. When the WPT system is in standby or operating mode, the detection system periodically starts a scan cycle at a frequency of 1Hz. The DSP controls the sweep frequency signal source to sequentially output excitation signals at 128 frequency points, and the current impedance spectrum is read through the impedance analysis module. Specifically, at each frequency point, the current Z-score can be read through the AD5941. b(ω) and θ(ω). Then, differential calculation and feature extraction are performed to calculate the differential impedance spectrum ΔZ(ω) of the current impedance spectrum and the reference spectrum, and four key feature vectors are extracted therefrom, including the maximum value in the ΔR(ω) spectrum and the corresponding frequency, the ΔR values at two characteristic frequency points of 100 kHz and 200 kHz, the ΔL values at two characteristic frequency points of 50 kHz and 150 kHz, and the sum of the modulus values of ΔZ(ω) in the entire frequency band as the overall change intensity; subsequently, the dynamic threshold calculation and intelligent discrimination stage is entered, the DSP obtains the current working current of the WPT system through the bus, dynamically adjusts the discrimination threshold of the key features according to the preset coefficient mapping table, and inputs all the extracted feature vectors into a pre-off-line trained classification model, which will output a specific discrimination result, such as 0-no abnormality, 1-copper, 2-aluminum, 3-steel, 4-stainless steel, etc. Finally, the system executes decision according to the discrimination result, if the result is no abnormality, the next detection cycle is continued, if the result is not zero, the audible and light alarms are triggered and the vehicle-mounted display screen prompts to check the forklift charging area, at the same time, the system executes emergency power reduction or stops charging operation, so as to realize strong anti-interference ability, low false alarm rate and excellent environmental robustness through multi-dimensional information perception, material identification and dynamic threshold adjustment.

[0088] It should be noted that the above classification model is preferably a pre-off-line trained machine learning model, such as a support vector machine (SVM) or a neural network. Specifically, the input of the model is the extracted four feature vectors, i.e. the maximum value of ΔR and the frequency, the characteristic frequency point ΔR value, the characteristic frequency point ΔL value and the overall change intensity, and the output is the foreign object type discrimination result, including 0: no abnormality, 1: copper, 2: aluminum, 3: steel, 4: stainless steel. The classification model can be obtained by training historical data, which can accurately distinguish the impedance spectrum fingerprints of different metal materials and reduce the false alarm rate.

[0089] The aforementioned live body detection unit, also known as LWD (Live Wireless Dective) sensing unit, can adopt a fusion solution of millimeter wave radar and infrared sensor. Millimeter wave radar is based on time-of-flight ranging principle, has the advantages of high ranging accuracy and strong environmental adaptability, can work stably in bad weather conditions, is not affected by changes in light intensity, and can effectively penetrate dust and non-metallic obstacles. While the infrared sensor is specifically used to detect the infrared radiation characteristics of living beings, with strong specificity and low false alarm rate. These two types of sensors can be arranged in four groups according to the upper, lower, left and right directions at the lower part of the electric forklift chassis, together forming a complete live body detection network. Correspondingly, the decision logic of the system can follow clear judgment rules: in one embodiment, when only the millimeter wave radar is triggered, the system can determine that it is in a safe state, and the indicator light can display green; when neither of the two sensors is triggered, the system can enter a high alert state, and the indicator light displays red; when only the infrared sensor is triggered, the system can enter a warning state, and the indicator light displays yellow; when both sensors are triggered, the system determines the highest risk level, and the indicator light displays red and requires the driver to intervene immediately.

[0090] In this embodiment, the decision logic of the specific live body detection unit can refer to the following table:

[0091] In addition, the temperature sensing unit is configured to continuously monitor the temperature changes of key parts including the charging coil and power electronics, to realize multiple protection functions, which can specifically be: providing overheat protection to prevent equipment damage and safety accidents; realizing temperature closed-loop control to automatically adjust the charging power when the detected temperature exceeds the safety threshold; warning of faults through temperature change trends to timely discover potential problems such as failure of the heat dissipation system or aging of components; optimizing charging parameters according to real-time temperature data to maintain optimal charging efficiency under the premise of safety.

[0092] The voltage and current sensing unit, on the other hand, is responsible for monitoring the electrical safety of the system, mainly for: real-time monitoring of input and output current to prevent overload, detecting voltage abnormal fluctuations to ensure the system works within the rated range, calculating real-time transmission power and providing feedback for power regulation by accurately measuring voltage and current values, identifying abnormal conditions such as short circuit and open circuit and detecting impedance mismatch problems, monitoring system efficiency by analyzing voltage and current phase difference, and serving as a key link in the safety interlocking mechanism, which can be linked with the control system to immediately cut off the power supply when dangerous parameters are detected.

[0093] The electromagnetic shielding unit is specifically used to suppress electromagnetic interference, which can effectively prevent external circuits from interfering with the wireless charging system, while limiting the impact of the system's own electromagnetic field on external devices, ensuring that the system meets electromagnetic compatibility requirements.

[0094] In summary, the entire safety protection module 6 works through multiple levels and multiple types of sensors to build a complete protection system from foreign object detection, living body identification to operation parameter monitoring, providing a solid safety guarantee for wireless charging systems.

[0095] In some embodiments, the omni-directional dynamic safety protection assembly also includes a safety protection module 14 of the vehicle-mounted receiving unit 8, which can preferably provide multiple protection mechanisms, including overcharge and over-discharge protection in cooperation with the battery management system 12, temperature monitoring of the receiving coil 9 and power elements, and comprehensive abnormal fault diagnosis functions. In addition to the display screen, the user indicator 15 can also be equipped with a three-color signal light to provide clear status indication and safety warning for the driver.

[0096] As shown in Figure 11 The entire electric forklift wireless charging process corresponding to the electric forklift wireless charging system starts with intelligent scheduling of the warehouse management system. The system actively issues charging instructions to specific forklifts according to the global operation state. This scheme deeply integrates charging demand and production scheduling, and realizes the transition from passive response to active planning. As the forklift approaches the designated charging area, the multi-modal positioning system is started. Through the design of dynamic coverage in different areas, different cameras cooperatively scan the groups of two-dimensional codes arranged on the path, effectively expanding the detection range and optimizing the scanning timing. These two-dimensional codes use unique three-dimensional encoding technology, embedding absolute coordinates, floor information, direction parameters, and environmental feature fingerprints, etc. multi-dimensional data, laying a rich information foundation for accurate positioning.

[0097] When the vehicle continues to approach, the system enters the fine positioning stage. By fusing the observation data of multiple cameras and running the multi-camera fusion algorithm, the pose calculation results are weighted and averaged with confidence weight pre-set, and abnormal fault tolerance processing is performed, finally achieving high docking accuracy. After accurate positioning is completed, the system immediately starts the multi-level safety protection mechanism. First, the impedance spectrum of the detection coil is obtained through wide-band scanning, and the unique frequency response fingerprint is used to identify the type of metal foreign objects. Then, combined with the fusion judgment logic of millimeter wave radar and infrared sensor, reliable detection of living body intrusion is realized.

[0098] Under the premise of confirming the safety of the environment, the energy transmission system is activated, and the optimal sub-coil combination in the activated array transmitting coil is dynamically selected based on the previous positioning result to compensate for the actual parking deviation. At the same time, the system continuously tracks the optimal efficiency operating point and dynamically adjusts the resonant frequency and output power to achieve adaptive and efficient charging throughout the whole process. During the whole charging process, the system real-time returns the charging progress and state data to the warehouse management system, and after the charging is completed, the management system updates the vehicle state and intelligently allocates subsequent work tasks, thereby forming a complete intelligent closed loop covering scheduling, positioning, protection, charging and rescheduling, and comprehensively improving the operation efficiency and energy efficiency management level of the forklift team.

[0099] In summary, the scheme of the present application has at least the following technical effects: 1. By adopting high-efficiency energy transmission technology, i.e., adopting the combination of DD / DDQ coils and S i C / G a N inverter, the energy transmission efficiency is improved. 2. By using multi-modal high-precision alignment, the fusion of visual and magnetic field sensors can achieve high-precision forklift wireless charging positioning. 3. By using multi-level FOD / LWD detection, intelligent safety protection is achieved. 4. By using the adaptive control algorithm of dynamic impedance matching and frequency tracking, the wireless charging process of the forklift is effectively optimized. 5. The wireless charging system of the electric forklift proposed in the present application meets the industrial design requirements, including IP67 protection, anti-vibration and dust, and an average trouble-free working time of more than 50,000 hours. 6. The system used in the present application is linked with WMS / FMS, which can effectively improve the scheduling efficiency.

[0100] The defects of the above scheme and the proposed solutions are the result of the inventors' careful research and practice, so the discovery process of the above problems and the solutions proposed by the present disclosure to solve the above problems should be the contribution of the inventors to the present disclosure in the process of the present disclosure.

[0101] It should be understood that, for the foregoing method embodiments, although the steps in the flowcharts are displayed in sequence according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise explicitly stated herein, there is no strict order limitation for the execution of these steps, and these steps can be executed in other orders. Moreover, at least part of the steps in the flowcharts of the method embodiments can include multiple sub-steps or multiple stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution order of these sub-steps or stages is not necessarily sequential, but can be executed in rotation or alternation with at least part of other steps or sub-steps or stages of other steps.

[0102] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage medium, and when executed, can include the processes of the above-mentioned embodiments of each method. Any reference to memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical storage, high-density embedded non-volatile memory, resistive memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration but not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc. The database involved in the embodiments provided in the present application can include at least one of a relational database and a non-relational database. The non-relational database can include a distributed database based on a block chain, etc., without being limited thereto. The processor involved in the embodiments provided in the present application can be a general-purpose processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., without being limited thereto.

[0103] The technical features of the above embodiments can be combined in any way. To make the description concise, not all possible combinations of the technical features in the above embodiments are described, but it should be considered that any combination of the technical features is within the scope of the present disclosure, as long as the combination does not exist contradictory. In the above embodiments, the description of each embodiment is focused on, and the parts not described in detail in one embodiment can be referred to the relevant description of other embodiments.

[0104] The terms "include", "includes" and "including" as used herein are meant to be analogous to "comprising", "comprises" and "comprising" and are used in their broadest sense. For example, a process, method, system, product or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements which have been expressly named. Other steps or elements can optionally be present and other steps or elements inherent in the process, method, system, product, or apparatus with the same effect.

[0105] As used herein, the term "plurality" means two or more. The term "and / or" describes associative relationship of associated objects, which means that there can be three relationships, for example, A and / or B, which can represent the following three cases: A exists alone, A and B exist together, and B exists alone. The character " / " generally represents that the associated objects before and after are in an "or" relationship.

[0106] As used herein, the terms "first" and "second" are merely to distinguish similar objects, and do not represent a specific order for the objects. Understandably, the "first" and "second" can be interchanged in a specific order or sequence as appropriate. It should be understood that the objects distinguished by "first" and "second" can be interchanged as appropriate, so that the embodiments described herein can be implemented in an order other than those illustrated or described herein.

[0107] The above-described embodiments only express several embodiments of the present application, which are described in detail and specifically, but should not be understood as a limitation on the scope of the patent. It should be noted that for those skilled in the art, without departing from the concept of the present application, a number of modifications and improvements can be made, which are within the scope of the present application. Therefore, the scope of protection of the patent of the present application should be subject to the appended claims.

Claims

1. A smart alignment method for wireless charging of electric forklifts, characterized in that, include: The system controls multiple image acquisition devices on the electric forklift to scan multiple positioning markers laid on the ground in the charging area and performs coarse positioning based on the scanning results. In response to the determination that the coarse positioning completion conditions are met, the attitude information of the electric forklift is determined based on the current positioning marker scanning results, and the fine positioning stage begins. In the fine positioning stage, based on the observation results of each image acquisition device on the corresponding positioning marker, the position coordinates of each image acquisition device in the world coordinate system are calculated. Based on the posture information and the preset positional relationship between each image acquisition device and the forklift center, the position coordinates of each image acquisition device are converted into the corresponding forklift center coordinates in the forklift coordinate system; Based on the position confidence level corresponding to each of the positioning identifiers, the calculated center coordinates of each forklift are fused to obtain the fused position coordinates of the electric forklift. Based on the fused position coordinates, the electric forklift is controlled to move until the conditions for precise positioning are met.

2. The method according to claim 1, characterized in that, The coarse positioning completion condition is: A specific image acquisition device successfully scanned a specific positioning marker, and the error between the electric forklift position calculated based on the positioning marker and the target point in the world coordinate system was less than the coarse positioning threshold.

3. The method according to claim 2, characterized in that, The specific image acquisition device is a rear image acquisition device, and the specific positioning mark is a positioning mark placed at the corresponding position of the rear of the forklift in the wireless charging area. And / or, the attitude information of the electric forklift is obtained based on the calculation results of the specific positioning mark by the specific image acquisition device.

4. The method according to claim 3, characterized in that, The plurality of image acquisition devices also include a left image acquisition device located on the left side of the electric forklift overhead guard and a right image acquisition device located on the right side of the overhead guard; The left image acquisition device corresponds to at least two positioning markers deployed on the left side of the wireless charging area; The right image acquisition device corresponds to at least two positioning markers deployed on the right side of the wireless charging area.

5. The method according to any one of claims 1 to 4, characterized in that, When there are multiple corresponding positioning markers for an image acquisition device, the step of calculating the position coordinates of the image acquisition device in the world coordinate system based on the observation results of the image acquisition device on the corresponding positioning markers includes: Based on the observation results of each positioning marker by the image acquisition device, the decoding information of each positioning marker is determined; Based on the decoding information of each positioning identifier, the candidate position coordinates of the image acquisition device are calculated respectively; Based on the position confidence of each candidate position coordinate and its corresponding positioning identifier, the position coordinates of the image acquisition device in the world coordinate system are determined.

6. The method according to claim 5, characterized in that, The step of determining the position coordinates of the image acquisition device in the world coordinate system based on the position confidence scores of each candidate position coordinate and its corresponding positioning identifier includes: Calculate the degree of difference between the coordinates of each candidate location; If the degree of difference is not greater than the consistency threshold, then the position coordinates of the image acquisition device in the world coordinate system are calculated by weighted averaging based on the position confidence corresponding to each candidate position coordinate. If the difference is greater than the consistency threshold, then the candidate location coordinates with the highest location confidence are determined as the location coordinates of the image acquisition device in the world coordinate system.

7. The method according to any one of claims 1 to 4, characterized in that, The step of fusing the calculated center coordinates of each forklift based on the position confidence level corresponding to each of the positioning identifiers to obtain the fused position coordinates of the electric forklift includes: Based on the position confidence of the positioning markers on which the center coordinates of each forklift are based, the fusion weight of the center coordinates of each forklift is calculated. The center coordinates of each forklift are weighted and averaged based on the fusion weights to calculate the fused position coordinates of the electric forklift.

8. The method according to claim 7, characterized in that, The step of calculating the fusion weight of the center coordinates of each forklift includes: Calculate the sum of the position confidence scores of the valid positioning identifiers corresponding to the center coordinates of the forklift; the valid positioning identifiers are those that are determined to meet the validity conditions when calculating the position coordinates of the image acquisition device. The fusion weight of the forklift center coordinates is obtained by calculating the ratio of the sum of the location confidence scores to the total number of valid location identifiers.

9. The method according to claim 8, characterized in that, The validity conditions of the location identifier include: The information that is successfully decoded and verified is matched with the environmental feature information detected in real time.

10. The method according to claim 8, characterized in that, The positioning identifier is a three-dimensional QR code, whose encoded information includes a basic positioning layer, a direction information layer, and an extended data layer; The basic positioning layer code includes the absolute coordinates of the positioning identifier in the world coordinate system, the floor identifier, and the check code. The direction information layer is encoded with the installation orientation information of the positioning identifier; The extended data layer encodes the location confidence level of the location identifier and the illumination level of the installation location.

11. The method according to any one of claims 1 to 4, characterized in that, After obtaining the fused position coordinates of the electric forklift, the method further includes: Calculate the degree of dispersion between the center coordinates of each forklift; If the degree of dispersion is greater than the variance threshold, then the sliding window filtering algorithm is enabled to smooth the fused position coordinates.

12. The method according to any one of claims 1 to 4, characterized in that, The conditions for achieving precise positioning include: The distance error between the fused location coordinates and the target charging location is no greater than the precision positioning threshold, and this state continues for a preset number of fused location coordinate calculation cycles.

13. A smart alignment assembly for wireless charging of electric forklifts, characterized in that, include: The spatial pose perception module includes multiple positioning markers deployed on the ground of the charging area and multiple image acquisition devices installed on the electric forklift. The decision control module is communicatively connected to the spatial pose perception module and is configured to perform the method as described in any one of claims 1 to 12.

14. A wireless charging system for electric forklifts, characterized in that, It includes a wireless charging component and a smart alignment component as described in claim 13, wherein the wireless charging component is communicatively connected to the smart alignment component.

Images