Miniaturized intelligent signal decoupling device suitable for piezoelectric multi-dimensional force measurement system

Abstract

The invention provides a miniaturized intelligent signal decoupling device suitable for a piezoelectric multi-dimensional force measurement system, and belongs to the technical field of signal measurement and decoupling. Comprising a plurality of piezoelectric six-dimensional force sensors which are used for collecting six-dimensional force original signals and outputting charge signals through cables; the front-end signal acquisition unit is used for receiving the charge signals output by the plurality of piezoelectric six-dimensional force sensors in parallel, processing the charge signals and outputting analog signals to the outside; the intelligent data decoupling unit is electrically connected with the front-end signal acquisition unit and is used for receiving the analog signal output by the front-end signal acquisition unit and performing data decoupling and measurement error compensation on the six-dimensional force original signal corresponding to the analog signal through a built-in decoupling and error compensation algorithm to obtain a decoupled six-dimensional force signal; wherein the intelligent data decoupling unit also provides a power supply for the front-end signal acquisition unit. The self-adaptive quasi-static charge leakage correction and temperature drift compensation algorithm is introduced, so that the decoupling precision is improved, and the size of equipment is simplified while signal crosstalk is eliminated through the stacked PCB structure.

Description

Technical Field

[0001] The present invention relates to the technical field of signal measurement and decoupling, and particularly to a miniaturized intelligent signal decoupling device suitable for a piezoelectric multi-dimensional force measurement system. Background Art

[0002] With the proposal of the concept of "embodied intelligence", the demand for multi-dimensional force sensors in the force feedback system of robots has gradually increased. Among them, piezoelectric F / T sensors have been widely used due to their technical advantages of strong structural stiffness, good measurement sensitivity and dynamic characteristics. In actual deployment, the charge signal generated by the piezoelectric F / T sensor needs to be first converted into a standard voltage signal through a charge amplification module, then converted into a digital signal through a signal acquisition card, and the measured F / T signal is obtained by data decoupling by the host computer.

[0003] In the application scenarios adapted to robots, the miniaturization, intelligence and expandability of the signal decoupling module of the piezoelectric six-dimensional force sensor are very important. At present, most of the signal decoupling devices suitable for piezoelectric six-dimensional force sensors adopt a design scheme of discrete modules. For example, the YE5853 multi-channel charge amplifier of Jiangsu Lianneng Electronics has a large volume, and a complex signal acquisition module and data decoupling module need to be adapted at the back end, and it cannot complete the signal decoupling work alone; the 5167A multi-channel charge measurement device of Kistler in Switzerland can complete the signal decoupling of single-dimensional force or three-dimensional force sensors, but when facing the measurement scenario equipped with a piezoelectric six-dimensional force sensor, the expandability of this device is weak.

[0004] Therefore, it is very necessary to provide a miniaturized intelligent signal decoupling device suitable for a piezoelectric multi-dimensional force measurement system to solve the problem of miniaturized hardware integration of multi-channel voltage signal acquisition equipment, not only reducing the volume of the signal acquisition equipment, but also improving the crosstalk problem of multi-channel signals, and improving the charge leakage and temperature drift of the measurement channels brought by the environment, providing error estimation and compensation, and improving the decoupling accuracy and long-term stability of the sensor. Summary of the Invention

[0005] In view of this, the present invention proposes a miniaturized intelligent signal decoupling device suitable for a piezoelectric multi-dimensional force measurement system, which introduces an adaptive quasi-static charge leakage correction and temperature drift compensation algorithm to improve the decoupling accuracy.

[0006] The present invention provides a miniaturized intelligent signal decoupling device suitable for a piezoelectric multi-dimensional force measurement system, including:

[0007] A plurality of piezoelectric six-dimensional force sensors, which collect six-dimensional force original signals and output charge signals through cables;

[0008] The front-end signal acquisition unit is used to receive the charge signals output by a number of piezoelectric six-axis force sensors in parallel, and after signal processing of the charge signals, output analog signals externally;

[0009] The intelligent data decoupling unit is electrically connected to the front-end signal acquisition unit, and is used to receive the analog signals output by the front-end signal acquisition unit, and through the built-in decoupling and error compensation algorithm, perform data decoupling and measurement error compensation on the six-axis force raw signals corresponding to the analog signals to obtain decoupled six-axis force signals;

[0010] Among them, the intelligent data decoupling unit also provides power to the front-end signal acquisition unit.

[0011] Based on the above technical solutions, preferably, the front-end signal acquisition unit includes a number of input channels and a number of charge signal processing modules. The number of charge signal processing modules are electrically connected to a number of piezoelectric six-axis force sensors one-to-one through the input channels, and are used to independently process the six-axis force raw signals received by each input channel.

[0012] Preferably, each of the number of charge signal processing modules includes a precision charge amplifier circuit, a programmable voltage gain circuit, an active low-pass filter circuit, a DC bias and reference voltage generation circuit, and an internal temperature compensation circuit; among them, the precision charge amplifier circuit is used to convert the charge signal corresponding to the input six-axis force raw signal into a primary voltage signal, and the converted primary voltage signal is sent into the programmable voltage gain circuit. The programmable voltage gain circuit converts the primary voltage signal to obtain a voltage signal with programmable gain; the active low-pass filter circuit receives the voltage signal with programmable gain, filters the voltage signal with programmable gain, and sends the denoised voltage signal into the DC bias and reference voltage generation circuit. The DC bias and reference voltage generation circuit adds a DC bias to the denoised voltage signal to obtain an analog signal that meets the input range of the intelligent data decoupling unit; the internal temperature compensation circuit respectively performs environmental temperature self-stabilization during the power-on process and constant temperature control during the on-line measurement process for each of the number of charge signal processing modules.

[0013] Further preferably, the precision charge amplification circuit includes an electrometer operational amplifier U1, a first resistor R1, a second resistor R2, a third resistor R3, a first capacitor C1, and a charge discharge analog switch SW1. One end of the first resistor R1 is electrically connected to the non-inverting input terminal of the electrometer operational amplifier U1, and the other end of the first resistor R1 is grounded. The inverting input terminal of the electrometer operational amplifier U1 is respectively electrically connected to one end of the second resistor R2, one end of the third resistor R3, one end of the first capacitor C1, and the normally open contact of the charge discharge analog switch SW1. The other end of the second resistor R2 is electrically connected to the output terminal of a piezoelectric six-axis force sensor through an input channel. The other ends of the third resistor R3, the other end of the first capacitor R1, and the normally open contact of the charge discharge analog switch SW1 are all electrically connected to the output terminal of the electrometer operational amplifier U1. After the coil of the charge discharge analog switch SW1 is attracted, the normally open contact of the charge discharge analog switch SW1 closes and is grounded. The precision charge amplification circuit converts the charge signal sent by the second resistor R2 into a primary voltage signal and outputs it to the programmable voltage gain circuit;

[0014] The programmable voltage gain circuit includes a programmable gain amplifier U2. The non-inverting input terminal of the programmable gain amplifier U2 is electrically connected to the output terminal of the electrometer operational amplifier U1. The inverting input terminal of the programmable gain amplifier U2 is grounded. The programmable terminal of the programmable gain amplifier U2 is communicatively connected to the intelligent data decoupling unit. The output terminal of the programmable gain amplifier U2 is electrically connected to the input terminal of the active low-pass filter circuit;

[0015] The active low-pass filter circuit includes two groups of series-connected low-pass filter modules. Each low-pass filter module includes an active amplifier U3, a fourth resistor R4, a fifth resistor R5, a second capacitor C2, and a third capacitor C3. The non-inverting input terminal of the active amplifier U3 is respectively electrically connected to one end of the fifth resistor R5 and one end of the third capacitor C3. The other end of the third capacitor C3 is grounded. The other end of the fifth resistor R5 is respectively electrically connected to one end of the second capacitor C2 and one end of the fourth resistor R4. The other end of the second capacitor C2 and the inverting input terminal of the active amplifier U3 are both electrically connected to the output terminal of the active amplifier U3. The other end of the fourth resistor R4 of the front-stage low-pass filter module is used as the input terminal of the active low-pass filter circuit and is electrically connected to the output terminal of the programmable gain amplifier U2. The output of the active amplifier U3 of the front-stage low-pass filter module is used as the input of the rear-stage low-pass filter module. The output of the active amplifier of the rear-stage low-pass filter module is used as the output of the active low-pass filter circuit;

[0016] The DC bias and reference voltage generation circuit includes a precision amplifier U4, a voltage reference chip U5, a sixth resistor R6, a seventh resistor R7, an eighth resistor R8, and a ninth resistor R9. The input end of the voltage reference chip U5 is electrically connected to an external power supply input. One end of the ninth resistor R9 is electrically connected to the output end of the voltage reference chip U5, and the other end of the ninth resistor R9 is respectively electrically connected to one end of the eighth resistor R8, the power supply input end of the programmable gain amplifier U2, the power supply input end of the internal temperature compensation circuit, and the input of the precision amplifier U4. The other end of the eighth resistor R8 is grounded; the input end of the precision amplifier U4 is also electrically connected to the output end of the active low-pass filter circuit. The inverting input end of the precision amplifier U4 is electrically connected to one end of the sixth resistor R6 and one end of the seventh resistor R7. The other end of the sixth resistor R6 is grounded, and the other end of the seventh resistor R7 is electrically connected to the output end of the precision amplifier U4. The output end of the precision amplifier U4 outputs an analog signal that meets the input range of the intelligent data decoupling unit;

[0017] The internal temperature compensation circuit includes a power driving unit and a heating power resistor array. The power driving unit adopts a totem-pole push-pull output structure based on NMOS transistors and is used to provide an input signal for the heating power resistor array; the heating power resistor array is arranged corresponding to a number of charge signal processing modules; the heating driving signal input end of the power driving unit is electrically connected to the intelligent data decoupling unit.

[0018] More preferably, the intelligent data decoupling unit includes a main controller, a human-machine interaction module, an on-board power supply, and a communication module; the main controller is connected to the human-machine interaction module and the communication module. The decoupling and error compensation algorithm built in the main controller is used to receive the analog signal output by the front-end signal acquisition unit, perform data decoupling and measurement error compensation on the analog signal, obtain the decoupled six-dimensional force signal, and send it to the human-machine interaction module and the communication module respectively; the on-board power supply supplies power to the main controller and the human-machine interaction module and serves as the external power supply input of a number of charge signal processing modules; the human-machine interaction module is used to output and display the decoupled six-dimensional force signal; the communication module is used for parameter configuration, device debugging, or real-time transmission of the decoupled six-dimensional force signal to the host computer or the Internet of Things device.

[0019] Even more preferably, the decoupling and measurement error compensation algorithm includes the following content:

[0020] S1: An ADC module is configured on the main controller, and the input channels of the ADC module are arranged corresponding to the output ends of a number of charge signal processing modules; the main controller adjusts the gain gear of the programmable voltage gain circuit to determine the range of the voltage signal after programmable gain;

[0021] S2: The main controller uses the M - order FIR band - pass filtering algorithm based on the Hanning window function to denoise the input analog signal, and obtains the original voltage signal U n and stores it in the FIFO buffer queue of the main controller. For the original voltage signal U that meets the voltage range of the analog signal n extract it, and compensate for the gain of the original voltage signal input to each input channel of the ADC module: U 1×n = G 1×n U n , n = 6, 7, 8, G 1×n is the gain compensation, and U is the original voltage signal after gain compensation 1×n , and then according to the charge mapping relationship of each channel, calculate the six - dimensional charge cumulative integral U linearly related to each force FT : U FT = [U Fx , U Fy , U Fz , U Mx , U Mx , U Mx = U 1×n I n×6 , U Fx , U Fy , U Fz , U Mx , U Mx , U Mx is the six - dimensional charge cumulative integral, and I n×6 is the channel mapping matrix;

[0022] S3: Use the piezoelectric conversion gain coefficient vector A 1×6 and the bias coefficient vector b 1×6 calibrated by the linear regression method: FT 1×6 = A 1×6 U FT + b 1×6 , FT 1×6 is the measured data of the six - dimensional force original signal after preliminary decoupling;

[0023] S4: Perform secondary decoupling on the measured data of the six - dimensional force original signal after preliminary decoupling, calibrate the multi - axis coupling error of the piezoelectric six - dimensional force sensor based on the least - squares method, and use the decoupling coefficient matrix obtained by the optimal estimation to solve the accurate measurement result FT d of the six - dimensional force original signal of the piezoelectric six - dimensional force sensor: FT d = D 6×6 FT 1×6 , where D 6×6 is the decoupling coefficient matrix, is the true measured value of the axial calibration test;

[0024] S5: The measurement error compensation algorithm includes quasi-static charge leakage correction and temperature drift compensation; the quasi-static charge leakage correction uses the charge discharge analog switch SW1 in the precision charge amplifier circuit. After the front-end signal acquisition unit completes initialization and starts the measurement task, a charge discharge reset operation is first performed, and the static charge signals U of several piezoelectric six-axis force sensors are continuously collected for a period of time Raw , the current ambient temperature T is recorded. When the amount of valid data measured reaches the set threshold, the charge leakage coefficients corresponding to several input channels are linearly fitted to obtain the charge leakage velocity v under the current ambient temperature parameter N and the quasi-static charge offset s N (T). The main controller calculates the real-time compensated charge measurement result U through the following formula N : U N = v N U Raw + s N (T), N = [1, 2,..., n];

[0025] The temperature drift compensation uses the internal temperature compensation circuit and the current ambient temperature T. After the front-end signal acquisition unit is powered on, according to the set temperature holding threshold, the output power of the heating power resistor array is adjusted by the proportional adjustment method to make the front-end signal acquisition unit reach the set operating temperature. The main controller adjusts the duty cycle of the heating drive signal of the power drive unit according to the deviation between the calculated temperature measurement value and the set temperature holding threshold, so that the operating temperature of the front-end signal acquisition unit remains constant;

[0026] S6: The main controller corrects according to the calculated real-time compensated charge measurement result, compensates for quasi-static charge leakage and temperature drift, and saves the compensation amount at the current temperature in the main controller.

[0027] Further preferably, in step S1, the main controller adjusts the gain level of the programmable voltage gain circuit to determine the range of the voltage signal after programmable gain. The range of the voltage signal after programmable gain is determined by the double-threshold method; first, the starting range of the voltage signal after programmable gain is set. The ADC module of the main controller collects the analog signal corresponding to the current piezoelectric six-axis force sensor output, and uses the sliding window method to filter the voltage data within the holding time range to obtain the average reference value of the voltage measurement result of the current analog signal. The upper and lower voltage threshold parameters and are set and compared with the measurement value. Let the upper voltage threshold parameter of the current range be V upper , the lower voltage threshold parameter of the current range be V lower , and the upper and lower voltage threshold parameters and be V th = V lower + V upper, the current measured value is V measured , the main controller determines the current optimal gain gear through the following discriminant formula: where δ is the allowable deviation amount, not exceeding 1% of the sum of the upper and lower threshold parameters of the current range voltage.

[0028] Further preferably, in step S2, the main controller uses the M-order FIR band-pass filtering algorithm based on the Hanning window function to denoise the input analog signal, including the following content: Define the impulse response h of the ideal low-pass filter LP (m): where f c is the cut-off frequency, f s is the sampling frequency, M is the filter order, m = 0, 1, 2,..., M, x is the variable in the parentheses; Let the lower cut-off frequency of the FIR band-pass be f1, and the upper cut-off frequency of the FIR band-pass be f2, then the impulse response function of the FIR band-pass is Multiply each point by the window function w(n) to suppress the frequency leakage in the FIR process: h(m) = h BP (m)·w(n), The frequency response function of the M-order FIR band-pass filtering algorithm based on the Hanning window function is

[0029] Further preferably, the on-board power supply obtains a 5V power rail through a DC regulated power supply, an adapter input or a USB interface. The 5V power is stepped down to obtain a 3.3V power rail for the main controller to use. The 5V power is stepped up and reversely processed to obtain +12V and -12V power rails for several charge signal processing modules to use.

[0030] Further preferably, it further includes a box body. The main controller and the on-board power supply of the intelligent data decoupling unit are integrated on the first PCB circuit board, and the first PCB circuit board is arranged at the middle position of the box body; several charge signal processing modules are respectively integrated on the second PCB circuit board and the third PCB circuit board. The second PCB circuit board and the third PCB circuit board are respectively and fixedly arranged at intervals on both sides of the first PCB circuit board in the box body. The human-computer interaction module and the communication module are embedded on different surfaces of the box body; several input channels are respectively arranged on the second PCB circuit board and the third PCB circuit board, and the input channels are used to be electrically connected to several piezoelectric six-dimensional force sensors in one-to-one correspondence; shielding covers are arranged on both sides of the second PCB circuit board and the third PCB circuit board; adjacent PCB circuit boards are fixed at intervals through copper posts.

[0031] A miniaturized intelligent signal decoupling device applicable to a piezoelectric multi-dimensional force measurement system provided by the present invention has the following beneficial effects compared with the prior art:

[0032] (1) The present invention proposes a miniaturized hardware integration solution for a multi-channel piezoelectric signal acquisition unit. It adopts a charge signal processing solution based on active feedback amplification, and through an up-and-down stacked PCB layout and an external shielding structure, while reducing the structural volume of the multi-channel piezoelectric signal acquisition unit, it improves the crosstalk problem of multi-channel signals.

[0033] (2) A software and hardware implementation solution for the quasi-static charge leakage correction and temperature drift compensation of a piezoelectric six-dimensional force sensor by the device in different environments. The miniaturization of the signal decoupling unit will inevitably affect the measurement accuracy and stability of the device. Therefore, it is necessary to additionally consider the charge leakage and temperature drift of the measurement channels brought by the environment. In the present invention, the intelligent data decoupling unit is equipped with an adaptive quasi-static charge leakage correction and temperature drift compensation algorithm. During the initialization process of the device, after obtaining the piezoelectric signals of the six-dimensional force sensor measured by each channel, the algorithm uses a digital FIR band-pass filter to denoise the original voltage signal, and performs F / T signal decoupling and error compensation on the multi-channel voltage signals output by the charge signal processing module, effectively improving the F / T decoupling accuracy and long-term stability of the piezoelectric six-dimensional force sensor. Description of the Drawings

[0034] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0035] Figure 1 It is the hardware structure block diagram of a miniaturized intelligent signal decoupling device applicable to a piezoelectric multi-dimensional force measurement system of the present invention;

[0036] Figure 2 It is the circuit schematic diagram of the charge signal processing module of a miniaturized intelligent signal decoupling device applicable to a piezoelectric multi-dimensional force measurement system of the present invention;

[0037] Figure 3 It is the structure block diagram of the intelligent data decoupling unit of a miniaturized intelligent signal decoupling device applicable to a piezoelectric multi-dimensional force measurement system of the present invention;

[0038] Figure 4 It is the structure block diagram of the on-board power supply of a miniaturized intelligent signal decoupling device applicable to a piezoelectric multi-dimensional force measurement system of the present invention;

[0039] Figure 5 It is the three-dimensional view of a miniaturized intelligent signal decoupling device applicable to a piezoelectric multi-dimensional force measurement system of the present invention after being boxed and encapsulated;

[0040] Figure 6 Schematic diagram of the PCB board stacking structure of a miniaturized intelligent signal decoupling device for a piezoelectric multi-dimensional force measurement system according to the present invention;

[0041] Figure 7 Flow chart of the program operation of a miniaturized intelligent signal decoupling device for a piezoelectric multi-dimensional force measurement system according to the present invention;

[0042] Figure 8 Flow chart of the quasi-static charge leakage correction and temperature drift compensation algorithm of a miniaturized intelligent signal decoupling device for a piezoelectric multi-dimensional force measurement system according to the present invention;

[0043] Figure 9 Flow chart of the multi-channel piezoelectric signal decoupling algorithm of a miniaturized intelligent signal decoupling device for a piezoelectric multi-dimensional force measurement system according to the present invention.

[0044] Reference numerals: 1, status indicator light; 2, LCD display screen; 3, upper housing of the box body; 4, USB-C communication interface; 5, CAN bus interface; 6, lower housing of the box body; 7, output interface of the piezoelectric six-dimensional force sensor; 8, second PCB circuit board; 9, third PCB circuit board; 10, first PCB circuit board; 11, XT30 DC power supply interface; 12, RS422 bus interface; 13, precision charge amplifier circuit, programmable voltage gain circuit and its shielding cover; 14, active low-pass filter circuit, DC bias and reference voltage generation circuit, internal temperature compensation circuit and its shielding cover; 15, bottom shielding cover; 16, direct plug-in signal connector between PCB boards. Detailed implementation manners

[0045] Next, in combination with the embodiments of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

[0046] The piezoelectric six-dimensional force sensor in the force feedback system of the robot is deployed in application scenarios such as industrial robotic arms, external force perception of humanoid robots, and force / torque (abbreviation: F / T) measurement and analysis of small actuators. However, the device has a large volume, weak scalability, and problems of charge leakage and temperature drift in the measurement channels. In view of this, as Figure 1 shown, the present invention provides a miniaturized intelligent signal decoupling device applicable to a piezoelectric multi-dimensional force measurement system, including:

[0047] A number of piezoelectric six-axis force sensors are used to collect six-axis force raw signals and output charge signals through cables; they can form 6-8 channels of charge signals, and the charge signals can be sent into the front-end signal acquisition unit through the SMA signal terminals of high-impedance cables.

[0048] The front-end signal acquisition unit is used to receive the charge signals output by a number of piezoelectric six-axis force sensors in parallel, and after signal processing of the charge signals, output analog signals outward;

[0049] The intelligent data decoupling unit is electrically connected to the front-end signal acquisition unit, and is used to receive the analog signals output by the front-end signal acquisition unit, perform ADC conversion on the analog signals, and through the built-in decoupling and error compensation algorithms, perform data decoupling and measurement error compensation on the six-axis force raw signals corresponding to the analog signals to obtain the decoupled six-axis force signals;

[0050] Among them, the intelligent data decoupling unit also provides power management, communication functions and control signal input to the front-end signal acquisition unit. The intelligent data decoupling unit transmits the obtained piezoelectric six-axis force to the host computer or Internet of Things device, and real-time feedbacks the measurement parameters and operating status of the device through the human-computer interaction interface.

[0051] As Figure 1 can be seen, this device integrates a multi-channel signal acquisition unit and an intelligent data decoupling unit, and can realize multi-channel piezoelectric signal measurement and F / T data decoupling of piezoelectric six-axis force sensors of various structures. In terms of hardware, this device is equipped with an 8-channel miniaturized intelligent data decoupling unit, and integrates a high-performance processor and various peripheral modules, and can realize functions such as voltage signal acquisition, six-axis force data decoupling and error compensation; at the same time, this device is equipped with a programmable digital communication interface and a human-computer interaction unit, which improves the expandability and usability of the signal decoupling device.

[0052] As Figure 1 shown, the front-end signal acquisition unit includes a number of input channels and a number of charge signal processing modules. The number of charge signal processing modules is electrically connected to a number of piezoelectric six-axis force sensors one by one through the input channels, and is used to independently process the six-axis force raw signals received by each input channel. The illustrated front-end signal acquisition unit has up to 8 input channels, namely CH1, CH2, ……, CH8. Here, the number of channels is only for illustration, and actually can be expanded to a larger number. The illustrated number of channels should not be regarded as a limitation to the technical solution.

[0053] Furthermore, as Figure 2As shown in the figure, several charge signal processing modules all include a precision charge amplifier circuit, a programmable voltage gain circuit, an active low-pass filter circuit, a DC bias and reference voltage generation circuit, and an internal temperature compensation circuit. Among them, the precision charge amplifier circuit is used to convert the charge signal corresponding to the input six-dimensional force raw signal into a primary voltage signal, and the converted primary voltage signal is sent into the programmable voltage gain circuit. The programmable voltage gain circuit converts the primary voltage signal to obtain a voltage signal with programmable gain. The active low-pass filter circuit receives the voltage signal with programmable gain, filters the voltage signal with programmable gain, and then sends the denoised voltage signal into the DC bias and reference voltage generation circuit. The DC bias and reference voltage generation circuit adds a DC bias to the denoised voltage signal to obtain an analog signal that meets the input range of the intelligent data decoupling unit. The internal temperature compensation circuit realizes the environmental temperature self-stabilization during the power-on process of several charge signal processing modules and the constant temperature control during the on-line measurement process respectively.

[0054] Combined with Figure 2 , the circuit structures and functions of the charge signal processing module will be described one by one:

[0055] 1) The precision charge amplifier circuit includes an electrometer operational amplifier U1, a first resistor R1, a second resistor R2, a third resistor R3, a first capacitor C1, and a charge discharge analog switch SW1. One end of the first resistor R1 is electrically connected to the non-inverting input terminal of the electrometer operational amplifier U1, and the other end of the first resistor R1 is grounded. The inverting input terminal of the electrometer operational amplifier U1 is respectively electrically connected to one end of the second resistor R2, one end of the third resistor R3, one end of the first capacitor C1, and the normally open contact of the charge discharge analog switch SW1. The other end of the second resistor R2 is electrically connected to the output terminal of a piezoelectric six-dimensional force sensor through an input channel. The other end of the third resistor R3, the other end of the first capacitor R1, and the normally open contact of the charge discharge analog switch SW1 are all electrically connected to the output terminal of the electrometer operational amplifier U1. After the coil of the charge discharge analog switch SW1 is attracted, the normally open contact of the charge discharge analog switch SW1 closes and is grounded. The precision charge amplifier circuit converts the charge signal sent by the second resistor R2 into a primary voltage signal and outputs it to the programmable voltage gain circuit.

[0056] In this embodiment, the electrometer operational amplifier U1 uses the ADA4530-1 of Analog Devices, Inc. Its leakage current is less than 20 fA, which can effectively reduce the input charge leakage under the quasi-static measurement of the device. The feedback loop of the electrometer operational amplifier U1 consists of a first resistor R1, a second resistor R2, a third resistor R3, and a first capacitor C1. The second resistor R2 is an input impedance matching resistor. The first resistor R1 and the first capacitor C1 form a charge integration loop. The first capacitor C1 is a C0G type ceramic capacitor with high insulation resistance and zero temperature drift coefficient, and its capacitance value is set to 100 nF. The third resistor R3 is a gain matching resistor used to improve the low-frequency characteristics of the electrometer operational amplifier U1, and its resistance value should be greater than 1 GΩ. The first resistor R1 is the input balance resistor of the electrometer operational amplifier U1, which is used to reduce the input bias current. The charge discharge analog switch SW1 uses a broadband TQ2SA type small signal relay, which can realize the active charge discharge function of the piezoelectric input channel through external control. In the PCB layout, the precision charge amplifier circuit arranges a signal ground loop protection ring on the input pins on both sides of the charge amplifier to reduce the coupling of environmental electrical noise. At the same time, the high-impedance charge input signal line is subjected to solder mask windowing to reduce the influence of bypass parasitic capacitance.

[0057] 2) The programmable voltage gain circuit includes a programmable gain amplifier U2. The non-inverting input terminal of the programmable gain amplifier U2 is electrically connected to the output terminal of the electrometer operational amplifier U1. The inverting input terminal of the programmable gain amplifier U2 is grounded. The programmable terminal of the programmable gain amplifier U2 is communicatively connected to the intelligent data decoupling unit. The output terminal of the programmable gain amplifier U2 is electrically connected to the input terminal of the active low-pass filter circuit.

[0058] In this embodiment, the programmable gain amplifier U2 uses the PGA113 of Texas Instruments. The binary gains of this amplifier include 1, 2, 4, 8, 16, 32, and 128, and the range gains include 1, 2, 5, 10, 20, 50, 100, and 200. The programmable gain amplifier U2 communicates with the intelligent data decoupling unit through its built-in SPI interface and can dynamically adjust the gain multiple.

[0059] 3) The active low-pass filter circuit includes two sets of series-connected low-pass filter modules. Each low-pass filter module includes an active amplifier U3, a fourth resistor R4, a fifth resistor R5, a second capacitor C2, and a third capacitor C3. The non-inverting input terminal of the active amplifier U3 is electrically connected to one end of the fifth resistor R5 and one end of the third capacitor C3 respectively. The other end of the third capacitor C3 is grounded. The other end of the fifth resistor R5 is electrically connected to one end of the second capacitor C2 and one end of the fourth resistor R4 respectively. The other end of the second capacitor C2 and the inverting input terminal of the active amplifier U3 are both electrically connected to the output terminal of the active amplifier U3. The other end of the fourth resistor R4 of the previous-stage low-pass filter module serves as the input terminal of the active low-pass filter circuit and is electrically connected to the output terminal of the programmable gain amplifier U2. The output of the active amplifier U3 of the previous-stage low-pass filter module serves as the input of the subsequent-stage low-pass filter module, and the output of the active amplifier of the subsequent-stage low-pass filter module serves as the output of the active low-pass filter circuit.

[0060] The low-pass filter module in this embodiment adopts a fourth-order Sallen-Key type low-pass filter topology structure. The active amplifier U3 uses an OPA1688 low-noise precision amplifier, and its cut-off frequency is determined by the impedance network composed of the second capacitor C2, the third capacitor C3, the fourth resistor R4, and the fifth resistor R5. The determined cut-off frequency in this embodiment is 2.5 kHz, which can maintain a good denoising effect while reducing the attenuation of the effective dynamic signal.

[0061] 4) The DC bias and reference voltage generation circuit includes a precision amplifier U4, a voltage reference chip U5, a sixth resistor R6, a seventh resistor R7, an eighth resistor R8, and a ninth resistor R9. The input terminal of the voltage reference chip U5 is electrically connected to the external power supply input. One end of the ninth resistor R9 is electrically connected to the output terminal of the voltage reference chip U5. The other end of the ninth resistor R9 is electrically connected to one end of the eighth resistor R8, the power supply input terminal of the programmable gain amplifier U2, the power supply input terminal of the internal temperature compensation circuit, and the input of the precision amplifier U4 respectively. The other end of the eighth resistor R8 is grounded; the input terminal of the precision amplifier U4 is also electrically connected to the output terminal of the active low-pass filter circuit. The inverting input terminal of the precision amplifier U4 is electrically connected to one end of the sixth resistor R6 and one end of the seventh resistor R7 respectively. The other end of the sixth resistor R6 is grounded. The other end of the seventh resistor R7 is electrically connected to the output terminal of the precision amplifier U4. The output terminal of the precision amplifier U4 outputs an analog signal that meets the input range of the intelligent data decoupling unit.

[0062] In this embodiment, the denoised voltage signal adds a stable DC bias signal through a DC bias and a reference voltage generation circuit, and outputs an analog signal that meets the input range of the intelligent data decoupling unit. Among them, the precision amplifier U4 is a non-inverting adder composed of GS8552. The sixth resistor R6 and the seventh resistor R7 are used to compensate for the fixed gain error of the voltage signal. The voltage reference chip U5 uses a REF5025 voltage reference chip that includes an ambient temperature signal output. The resistance values of the eighth resistor R8 and the ninth resistor R9 are set to 2 kΩ and 3.3 kΩ respectively, and a reference voltage of about 1.55 V is output, which is used to provide a DC bias voltage input with high PSRR for the programmable gain amplifier U2 and the precision amplifier U4 to meet the static reference point of the operational amplifier under the condition of single power supply.

[0063] 5) The internal temperature compensation circuit includes a power driving unit and a heating power resistor array. The power driving unit adopts a totem pole push-pull output structure based on NMOS transistors, which is used to provide an input signal for the heating power resistor array. The heating power resistor array is correspondingly arranged at several charge signal processing modules. The heating drive signal input terminal of the power driving unit is electrically connected to the intelligent data decoupling unit. The totem pole push-pull output structure based on NMOS transistors is a conventional technical means in the art and will not be elaborated here. The heating power resistor array is formed by series / parallel connection of precision resistors with good consistency, and the heat generated by the resistors arranged in this way is relatively uniform.

[0064] As Figure 3 shown, the intelligent data decoupling unit includes a main controller, a human-machine interaction module, an on-board power supply, and a communication module. The main controller is connected to the human-machine interaction module and the communication module. The decoupling and error compensation algorithm built into the main controller is used to receive the analog signal output by the front-end signal acquisition unit, perform data decoupling and measurement error compensation on the analog signal, obtain the decoupled six-dimensional force signal, and send it to the human-machine interaction module and the communication module respectively. The on-board power supply supplies power to the main controller and the human-machine interaction module, and serves as the external power input of several charge signal processing modules. The human-machine interaction module is used to output and display the decoupled six-dimensional force signal. The communication module is used for parameter configuration, device debugging, or real-time transmission of the decoupled six-dimensional force signal to the host computer or Internet of Things device.

[0065] In this embodiment, the main controller adopts a high-performance 32-bit STM32H743 microcontroller chip, whose operating frequency is 480 MHz and supports a DSP operation library with double-precision floating-point, capable of meeting the computing power requirements for real-time deployment of multi-channel charge data decoupling and error compensation algorithms. At the same time, it contains a high-speed SAR-type ADC module with 16-bit precision, supports hardware oversampling and DMA transfer functions, and is equipped with peripheral interfaces for data communication buses such as SPI, I2C, and USART, capable of supporting complex data and control bus connections in the front-end signal acquisition unit. In terms of storage peripherals, the main controller is respectively equipped with 32 MB of RAM and ROM memory. The RAM uses the DDR chip IS42S32800, which is connected to the STM32 microcontroller through the FMC parallel bus; the ROM memory uses the NOR Flash chip W25Q256, which is connected to the microcontroller through the QSPI bus.

[0066] The intelligent data decoupling unit has a total of 3 independent external serial communication interfaces. Among them, the CAN interface and the USB-C interface are respectively connected to the PHY unit inside the main controller. The RS422 interface uses the SIT3088E chip for interface level conversion and is connected to the USART_TX / RX peripheral interface of the main controller; in terms of function, the USB-C communication interface is dedicated to parameter configuration and firmware debugging of the main controller, while the RS422 bus interface and the CAN bus interface are used for initialization configuration of the intelligent data decoupling unit and real-time transmission of decoupled data of the six-axis force sensor. The human-machine interaction module consists of 2 groups of LED status indicators, a 2.5-inch LCD display, and a buzzer, which are automatically controlled by the main controller through the GPIO and I2C / SPI buses. The on-board power supply powers the front-end signal acquisition unit through a signal connector. The analog multi-channel raw voltage and temperature measurement signals output by the front-end signal acquisition unit are input to the internal ADC module pins of the main controller, and the REF3030 voltage reference chip is used to provide a low-noise voltage reference for the ADC; its charge discharge control signal and heating PWM drive signal are provided by the GPIO digital interface of the main controller, while the voltage gain status signal is polled and read by setting the chip select (CS) level through the SPI bus, and the main controller performs dual-threshold automatic gain control to select an appropriate gain multiple for the programmable gain amplifier U2.

[0067] As Figure 4 shown, the on-board power supply obtains a 5V power rail through a DC regulated power supply, an adapter input, or a USB interface. The 5V power is stepped down to obtain a 3.3V power rail for the main controller to use. The 5V power is stepped up and inverted to obtain +12V and -12V power rails for several charge signal processing modules to use.

[0068] In terms of design, it supports two different input power supplies. One is a regulated DC power supply or adapter input, whose input voltage supports a wide range of 6 - 16V, and uses a TPS54560 switching buck chip to output a 5V power rail for powering the main system of the device; the other is a standard USB 5V@1A input, which uses a TPD3S044 chip for ESD protection and overcurrent protection of the USB port and inputs a 5V voltage rail. Two parallel 5V power branches are converged by connecting Schottky diodes in series to prevent power backflow. After that, the 5V power rail outputs a 3.3V voltage rail through a SY8089 switching buck chip for powering the main controller and its peripherals; at the same time, the bypass outputs a 15V DC voltage through an LM27313 switching boost chip and uses an ICL7662 inverting charge pump chip to output a -15V DC voltage. Finally, +12V and -12V power rails are output through a low-noise LM317 / 337 linear voltage regulator chipset for powering the analog path of the front-end signal acquisition unit, thus meeting the power supply requirements of chips with different voltage levels.

[0069] The decoupling and error compensation algorithms built into the main controller include the following:

[0070] S1: As Figure 9 shown, the flow of the decoupling algorithm is as shown in the figure. The main controller is configured with an ADC module, and the input channels of the ADC module are set in one-to-one correspondence with the output ends of several charge signal processing modules; the main controller adjusts the gain level of the programmable voltage gain circuit to determine the range of the voltage signal after programmable gain.

[0071] In step S1, when the main controller adjusts the gain level of the programmable voltage gain circuit to determine the range of the voltage signal after programmable gain, it determines the range of the voltage signal after programmable gain through the dual-threshold method; first, set the starting range of the voltage signal after programmable gain. The ADC module of the main controller collects the analog signal corresponding to the current piezoelectric six-axis force sensor output, filters the voltage data within the hold time range using the sliding window method to obtain the average reference value of the voltage measurement result of the current analog signal, sets the upper and lower voltage threshold parameters V upper , the lower voltage threshold parameter of the current range is V lower , and the sum of the upper and lower voltage threshold parameters is V th = V lower + V upper , the current measured value is V measured , and the main controller determines the current optimal gain level through the following discriminant: where δ is the allowable deviation amount, not exceeding 1% of the sum of the upper and lower voltage threshold parameters of the current range.

[0072] S2: The main controller uses the M - order FIR band - pass filtering algorithm based on the Hanning window function to denoise the input analog signal, and obtains the original voltage signal U n and stores it in the FIFO buffer queue of the main controller. For the original voltage signal U that meets the voltage range of the analog signal n extract it, and compensate for the gain of the original voltage signal input to each input channel of the ADC module: U 1×n = G 1×n U n , n = 6, 7, 8, G l×n is the gain compensation, and U l×n is the original voltage signal after compensation gain. Then, according to the charge mapping relationship of each channel, calculate the six - dimensional charge cumulative integral U FT related to the linear force in each direction: U FT = [U Fx , U Fy , U Fz , U Mx , U Mx , U Mx = U l×n I n×6 , U Fx , U Fy , U Fz , U Mx , U Mx , U Mx is the six - dimensional charge cumulative integral, and I n×6 is the channel mapping matrix.

[0073] In step S2, the main controller uses the M - order FIR band - pass filtering algorithm based on the Hanning window function to denoise the input analog signal, including the following content: Define the impulse response h LP (m) of the ideal low - pass filter: where f c is the cut - off frequency, f s is the sampling frequency, M is the filter order, m = 0, 1, 2,..., M, x is the variable in the parentheses; Let the lower cut - off frequency of the FIR band - pass be f1, and the upper cut - off frequency of the FIR band - pass be f2, then the impulse response function of the FIR band - pass is Multiply each point by the window function w(n) to suppress frequency leakage in the FIR process: h(m) = h BP (m)·w(n), The frequency response function of the M - order FIR band - pass filtering algorithm based on the Hanning window function is

[0074] S3: Use the piezoelectric conversion gain coefficient vector A calibrated by the linear regression method for the piezoelectric six - dimensional force sensor1×6 and the bias coefficient vector b 1×6 : FT 1×6 = A 1×6 U FT + b 1×6 , where FT 1×6 is the measured data of the initially decoupled six - dimensional force raw signal.

[0075] S4: Perform secondary decoupling on the measured data of the initially decoupled six - dimensional force raw signal, calibrate the multi - axis coupling error of the piezoelectric six - dimensional force sensor based on the least - squares method, and use the decoupling coefficient matrix obtained by the optimal estimation to solve the accurate measurement result FT of the six - dimensional force raw signal of the piezoelectric six - dimensional force sensor d : FT d = D 6×6 FT 1×6 , where D 6×6 is the decoupling coefficient matrix, is the true measured value of the axial calibration test.

[0076] S5: As shown in Figure 8 , the measurement error compensation algorithm includes quasi - static charge leakage correction and temperature drift compensation; the quasi - static charge leakage correction is to use the charge discharge analog switch SW1 in the precision charge amplifier circuit. After the front - end signal acquisition unit completes initialization and starts the measurement task, first perform a charge discharge reset operation, continuously collect the static charge signals U Raw of several piezoelectric six - dimensional force sensors for a period of time, record the current ambient temperature T. When the amount of valid measured data reaches the set threshold, linearly fit the charge leakage coefficients corresponding to several input channels to obtain the charge leakage speed v N and the quasi - static charge offset s N (T) at the current ambient temperature parameter. The main controller calculates the real - time compensated charge measurement result U N through the following formula: U N = v N U Raw + s N (T), N = [1, 2,..., n];

[0077] The temperature drift compensation is to use the internal temperature compensation circuit and the current ambient temperature T. After the front - end signal acquisition unit is powered on, according to the set temperature holding threshold, adjust the output power of the heating power resistor array through the proportional regulation method to make the front - end signal acquisition unit reach the set operating temperature. The main controller adjusts the duty cycle of the heating drive signal of the power drive unit according to the deviation between the calculated temperature measurement value and the set temperature holding threshold, so that the operating temperature of the front - end signal acquisition unit remains constant.

[0078] S6: The main controller corrects according to the calculated real-time compensation charge measurement result, compensates for quasi-static charge leakage and temperature drift, and stores the compensation amount at the current temperature in the main controller.

[0079] The main program running process of the intelligent data decoupling unit is as Figure 7 shown. After the device is powered on and started, the data decoupling unit starts to initialize, waits for the input signals of each channel of the sensor to be stable, and completes the quasi-static charge leakage correction and temperature drift compensation of the original piezoelectric signal. After that, the system will wait for the measurement task start instruction received by the RS422 / CAN communication interface through the serial port interrupt mechanism; if there is no response within the waiting time, the system will automatically enter the low-power standby state and can be woken up by an external communication bus event. After the device is started, the ADC module of the intelligent signal decoupling unit will collect the piezoelectric signals of the effective input channels, adaptively adjust the gain gear of the PGA113 according to the dynamic range of the input voltage of the charge amplifier, and determine the voltage range of each channel by the method of double thresholds. After that, the main controller uses the M-order FIR band-pass filtering algorithm based on the Hann window function to denoise the original voltage signals of multiple channels, and solves the F / T measurement values of the sensor through the six-dimensional force signal decoupling algorithm. Finally, the system displays the measurement status of each F / T sensor through the status indication module, and outputs the measurement results of the sensor in real time through the serial communication interface, and waits to repeat the signal decoupling process after the current polling cycle is completed. If the measurement task ends, the system will enter the standby state again and continue to wait for the next measurement task. In this embodiment, the order M takes a value of 50.

[0080] As Figure 5 and Figure 6 shown, in order to reduce the volume of the device, the present invention also provides an integrated hardware solution. The present invention also includes a box body. The main controller of the intelligent data decoupling unit and the on-board power supply are integrated on the first PCB circuit board, and the first PCB circuit board is arranged at the middle position of the box body; several charge signal processing modules are respectively integrated on the second PCB circuit board and the third PCB circuit board, and the second PCB circuit board and the third PCB circuit board are respectively fixedly arranged at intervals on both sides of the first PCB circuit board in the box body, and the human-computer interaction module and the communication module are embedded on different surfaces of the box body; several input channels are respectively arranged on the second PCB circuit board and the third PCB circuit board, and the input channels are used for corresponding electrical connection with several piezoelectric six-dimensional force sensors one by one; shielding covers are arranged on both sides of the second PCB circuit board and the third PCB circuit board; adjacent PCB circuit boards are fixedly spaced by copper posts.

[0081] As Figure 5 and Figure 6As shown, the overall packaging volume of the device is L*W*H = 80*60*50 mm. The box body is divided into an aluminum upper shell 3 and a lower shell 6, and the upper shell 3 covers the lower shell 6. The inside of the box body is hollow. The status indicator light 1 and the LCD display screen 2 are arranged on one end face of the upper cover body 3; the USB-C communication interface 4 and the CAN bus interface 5 are embedded on one side face of the lower shell 6, and the XT30 DC power supply interface 11 and the RS422 bus interface 12 are embedded on the other side face of the lower shell 6. The output interface 7 of the piezoelectric six-axis force sensor, that is Figure 1 The SMA terminals in are respectively electrically connected in one-to-one correspondence with different charge signal processing modules on the second PCB circuit board 8 and the third PCB circuit board 9, and are electrically connected to the piezoelectric six-axis force sensor through high-impedance cables. On the side of the second PCB circuit board 8 and the third PCB circuit board 9 away from the first PCB circuit board 10, there are arranged a precision charge amplification circuit, a programmable voltage gain circuit and its shielding cover (corresponding mark 13), as well as an active low-pass filter circuit, a DC bias and reference voltage generation circuit, an internal temperature compensation circuit and its shielding cover (corresponding mark 14). On the side of the second PCB circuit board 8 and the third PCB circuit board 9 close to the first PCB circuit board 10, there is a bottom shielding cover 15.

[0082] The internal circuit board connection method of this device adopts a 3-layer stacked PCB structure. Among them, the front-end signal acquisition unit is divided into 2 groups of the same PCBs, which are respectively located on the upper and lower layers of the structure, and are fixedly connected to the first PCB circuit board where the intermediate intelligent data decoupling unit is located through the direct plug signal connector 16 between the PCB boards. A copper shielding cover is arranged on the inner side of the PCB circuit board where the front-end signal acquisition unit is located to isolate the high-frequency digital noise of the intermediate layer data processing unit; the outer side includes the circuit area of the charge signal processing module, and 3 groups of separated copper shielding covers are respectively arranged to shield the electrostatic noise coupled by the environment. The stacked PCB circuit boards are fixed between the boards using copper posts and are installed in the internal positioning threaded holes of the aluminum upper / lower shells through bolts.

[0083] The above is only the preferred embodiment of the present invention and is not intended to limit the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A miniaturized intelligent signal decoupling device suitable for a piezoelectric multi-dimensional force measurement system, characterized in that: include: Several piezoelectric six-dimensional force sensors collect six-dimensional force raw signals and output charge signals through cables; The front-end signal acquisition unit is used to receive the charge signals output by a plurality of piezoelectric six-dimensional force sensors in parallel, and output analog signals after processing the charge signals; An intelligent data decoupling unit is electrically connected to the front-end signal acquisition unit, and is used to receive the analog signal output by the front-end signal acquisition unit, and perform data decoupling and measurement error compensation on the six-dimensional force original signal corresponding to the analog signal through a built-in decoupling and error compensation algorithm to obtain a decoupled six-dimensional force signal; Among them, the intelligent data decoupling unit also provides power to the front-end signal acquisition unit.

2. A miniaturized intelligent signal decoupling device suitable for a piezoelectric multi-dimensional force measurement system according to claim 1, characterized in that: The front-end signal acquisition unit includes a plurality of input channels and a plurality of charge signal processing modules. The plurality of charge signal processing modules are electrically connected to a plurality of piezoelectric six-dimensional force sensors through the input channels in a one-to-one correspondence, and are used to independently process the six-dimensional force original signal received by each input channel.

3. A miniaturized intelligent signal decoupling device suitable for a piezoelectric multi-dimensional force measurement system according to claim 2, characterized in that: Several charge signal processing modules all include a precision charge amplifier circuit, a programmable voltage gain circuit, an active low-pass filter circuit, a DC bias and reference voltage generating circuit and an internal temperature compensation circuit; wherein, the precision charge amplifier circuit is used to convert the charge signal corresponding to the input six-dimensional force original signal into a primary voltage signal, and the converted primary voltage signal is sent to the programmable voltage gain circuit, and the programmable voltage gain circuit converts the primary voltage signal to obtain a voltage signal after programmable gain; the active low-pass filter circuit receives the voltage signal after programmable gain, filters the voltage signal after programmable gain, and obtains a denoised voltage signal, which is sent to the DC bias and reference voltage generating circuit, and the DC bias and reference voltage generating circuit adds a DC bias to the denoised voltage signal to obtain an analog signal that meets the input range of the intelligent data decoupling unit; the internal temperature compensation circuits are respectively for the ambient temperature self-stabilization during the power-on process of several charge signal processing modules and the constant temperature control during the online measurement process.

4. A miniaturized intelligent signal decoupling device suitable for a piezoelectric multi-dimensional force measurement system according to claim 3, characterized in that: The precision charge amplifier circuit includes an electrometer operational amplifier U1, a first resistor R1, a second resistor R2, a third resistor R3, a first capacitor C1, and a charge discharge analog switch SW1. One end of the first resistor R1 is electrically connected to the in-phase input end of the electrometer operational amplifier U1, and the other end of the first resistor R1 is grounded. The inverting input end of the electrometer operational amplifier U1 is electrically connected to one end of the second resistor R2, one end of the third resistor R3, one end of the first capacitor C1, and a normally open contact of the charge discharge analog switch SW1, respectively. The other end of the second resistor R2 is electrically connected to the output end of a piezoelectric six-dimensional force sensor through an input channel. The other end of the third resistor R3, the other end of the first capacitor R1, and the normally open contact of the charge discharge analog switch SW1 are all electrically connected to the output end of the electrometer operational amplifier U1. After the coil of the charge discharge analog switch SW1 is attracted, the normally open contact of the charge discharge analog switch SW1 is closed and grounded. The precision charge amplifier circuit converts the charge signal sent by the second resistor R2 into a primary voltage signal and outputs it to the programmable voltage gain circuit. The programmable voltage gain circuit includes a programmable gain amplifier U2, the in-phase input terminal of the programmable gain amplifier U2 is electrically connected to the output terminal of the electrometer operational amplifier U1, the inverting input terminal of the programmable gain amplifier U2 is grounded, and the programmable terminal of the programmable gain amplifier U2 is communicatively connected to the intelligent data decoupling unit; the output terminal of the programmable gain amplifier U2 is electrically connected to the input terminal of the active low-pass filter circuit; The active low-pass filter circuit includes two groups of low-pass filter modules connected in series, each of which includes an active amplifier U3, a fourth resistor R4, a fifth resistor R5, a second capacitor C2 and a third capacitor C3. The in-phase input terminal of the active amplifier U3 is electrically connected to one end of the fifth resistor R5 and one end of the third capacitor C3 respectively, the other end of the third capacitor C3 is grounded, the other end of the fifth resistor R5 is electrically connected to one end of the second capacitor C2 and one end of the fourth resistor R4 respectively, the other end of the second capacitor C2 and the inverting input terminal of the active amplifier U3 are both electrically connected to the output terminal of the active amplifier U3, the other end of the fourth resistor R4 of the front-stage low-pass filter module is electrically connected to the output terminal of the programmable gain amplifier U2 as the input terminal of the active low-pass filter circuit, the output of the active amplifier U3 of the front-stage low-pass filter module is used as the input of the rear-stage low-pass filter module, and the output of the active amplifier of the rear-stage low-pass filter module is used as the output of the active low-pass filter circuit; The DC bias and reference voltage generating circuit includes a precision amplifier U4, a voltage reference chip U5, a sixth resistor R6, a seventh resistor R7, an eighth resistor R8 and a ninth resistor R9. The input end of the voltage reference chip U5 is electrically connected to the external power input, one end of the ninth resistor R9 is electrically connected to the output end of the voltage reference chip U5, the other end of the ninth resistor R9 is electrically connected to one end of the eighth resistor R8, the power input end of the programmable gain amplifier U2, the power input end of the internal temperature compensation circuit and the input of the precision amplifier U4, respectively, and the other end of the eighth resistor R8 is grounded; the input end of the precision amplifier U4 is also electrically connected to the output end of the active low-pass filter circuit, the inverting input end of the precision amplifier U4 is electrically connected to one end of the sixth resistor R6 and one end of the seventh resistor R7, the other end of the sixth resistor R6 is grounded, the other end of the seventh resistor R7 is electrically connected to the output end of the precision amplifier U4, and the output end of the precision amplifier U4 outputs an analog signal that meets the input range of the intelligent data decoupling unit; The internal temperature compensation circuit includes a power driving unit and a heating power resistor array. The power driving unit adopts a totem pole push-pull output structure based on an NMOS tube to provide an input signal for the heating power resistor array; the heating power resistor array is arranged one by one at a number of charge signal processing modules; the heating driving signal input end of the power driving unit is electrically connected to the intelligent data decoupling unit.

5. A miniaturized intelligent signal decoupling device suitable for a piezoelectric multi-dimensional force measurement system according to claim 4, characterized in that: The intelligent data decoupling unit includes a main controller, a human-computer interaction module, an onboard power supply and a communication module; the main controller is connected to the human-computer interaction module and the communication module, and the main controller has a built-in decoupling and error compensation algorithm, which is used to receive the analog signal output by the front-end signal acquisition unit, perform data decoupling and measurement error compensation on the analog signal, obtain the decoupled six-dimensional force signal and send it to the human-computer interaction module and the communication module respectively; the onboard power supply supplies power to the main controller and the human-computer interaction module, and serves as an external power input for several charge signal processing modules; the human-computer interaction module is used to output and display the decoupled six-dimensional force signal; the communication module is used for parameter configuration, equipment debugging or real-time transmission of the decoupled six-dimensional force signal to a host computer or an Internet of Things device.

6. A miniaturized intelligent signal decoupling device suitable for a piezoelectric multi-dimensional force measurement system according to claim 5, characterized in that: Decoupling and measurement error compensation algorithms, including the following: S1: The main controller is equipped with an ADC module, and the input channels of the ADC module are set in one-to-one correspondence with the output ends of several charge signal processing modules; the main controller adjusts the gain level of the programmable voltage gain circuit to determine the range of the voltage signal after the programmable gain; S2: The main controller uses the M-order FIR bandpass filter algorithm based on the Hanning window function to denoise the input analog signal and obtain the original voltage signal U n The original voltage signal U that meets the voltage range of the analog signal is stored in the FIFO buffer queue of the main controller. n Extract and compensate the gain of the original voltage signal input by each input channel of the ADC module: U 1×n =G 1×n U n , n=6,7,8,G 1×n is the gain compensation, and is the original voltage signal U after the compensation gain 1×n Then, according to the charge mapping relationship of each channel, the six-dimensional charge accumulation integral U linearly related to each force is calculated. FT :U FT =[U Fx , U Fy , ,U Fz , U Mx , U Mx , U Mx ]=U 1×n I n×6 , U Fx , U Fy ,U Fz ,U Mx , U Mx , U Mx is the six-dimensional charge accumulation integral, I n×6 is the channel mapping matrix; S3: Piezoelectric conversion gain coefficient vector A of the piezoelectric six-axis force sensor calibrated using the linear regression method 1×6 and the bias coefficient vector b 1×6 : FT 1×6 =A 1×6 U FT +b 1×6 , FT 1×6 The six-dimensional force raw signal measurement data for preliminary decoupling; S4: Perform secondary decoupling on the initially decoupled six-axis force original signal measurement data, calibrate the multi-axis coupling error of the piezoelectric six-axis force sensor based on the least squares method, and use the decoupling coefficient matrix obtained by optimal estimation Solve the precise measurement result FT of the original six-dimensional force signal of the piezoelectric six-dimensional force sensor d : FT d =D 6×6 FT 1×6 ,in D 6×6 is the decoupling coefficient matrix, is the true measurement value of the axial calibration test; S5: The measurement error compensation algorithm includes quasi-static charge leakage correction and temperature drift compensation; the quasi-static charge leakage correction uses the charge discharge analog switch SW1 in the precision charge amplifier circuit. After the front-end signal acquisition unit completes the initialization and starts the measurement task, it first performs a charge discharge reset operation and continuously collects the static charge signals U of several piezoelectric six-axis force sensors for a period of time. Raw , record the current ambient temperature T, and when the effective data volume measured reaches the set threshold, perform linear fitting on the charge leakage coefficients corresponding to several input channels to obtain the charge leakage rate v under the current ambient temperature parameters N and the quasi-static charge bias s N (T), the main controller calculates the real-time compensation charge measurement result U through the following formula N :U N =v N U Raw +s N (T), N = [1, 2, ..., n]; Temperature drift compensation uses the internal temperature compensation circuit and the current ambient temperature T. After the front-end signal acquisition unit is powered on, the output power of the heating power resistor array is adjusted by the proportional adjustment method according to the set temperature holding threshold so that the front-end signal acquisition unit reaches the set working temperature. The main controller adjusts the duty cycle of the heating drive signal of the power drive unit according to the deviation between the calculated temperature measurement value and the set temperature holding threshold so that the working temperature of the front-end signal acquisition unit remains constant. S6: The main controller corrects the calculated real-time compensation charge measurement result, compensates for the quasi-static charge leakage and temperature drift, and saves the compensation amount at the current temperature in the main controller.

7. A miniaturized intelligent signal decoupling device suitable for a piezoelectric multi-dimensional force measurement system according to claim 6, characterized in that: The main controller described in step S1 adjusts the gain gear of the programmable voltage gain circuit to determine the range of the voltage signal after the programmable gain, and determines the range of the voltage signal after the programmable gain by a double threshold method; first, the starting range of the voltage signal after the programmable gain is set, and the ADC module of the main controller collects the analog signal corresponding to the output of the current piezoelectric six-dimensional force sensor, and uses the sliding window method to filter the voltage data within the retention time range to obtain the average reference value of the voltage measurement result of the current analog signal, and sets the voltage upper and lower threshold parameters and, and compares them with the measured value, and the voltage upper threshold parameter of the current range is set to V upper , the voltage lower threshold parameter of the current range is V lower The voltage upper and lower threshold parameters are V th =V lower +V upper , the current measured value is V measured , the main controller determines the current optimal gain gear through the following discriminant: Where δ is the allowable deviation, which shall not exceed 1% of the sum of the upper and lower threshold parameters of the current range voltage.

8. A miniaturized intelligent signal decoupling device suitable for a piezoelectric multi-dimensional force measurement system according to claim 6, characterized in that: The main controller described in step S2 uses an M-order FIR bandpass filtering algorithm based on a Hanning window function to denoise the input analog signal, including the following contents: defining the impulse response h of an ideal low-pass filter LP (m): where f c is the cut-off frequency, f s is the sampling frequency, M is the filter order, m=0,1,2,...,M, x is the variable in brackets; let the lower cutoff frequency of FIR bandpass be f1, and the upper cutoff frequency of FIR bandpass be f2, then the impulse response function of FIR bandpass is Multiply each point by the window function w(n) to suppress frequency leakage in the FIR process: h(m) = h BP (m)·w(n), The frequency response function of the M-order FIR bandpass filter algorithm based on the Hanning window function is:

9. A miniaturized intelligent signal decoupling device suitable for a piezoelectric multi-dimensional force measurement system according to claim 5, characterized in that: The onboard power supply obtains a 5V power rail through a DC regulated power supply, adapter input or USB interface. The 5V power supply is stepped down to obtain a 3.3V power rail for the main controller. The 5V power supply is stepped up and reversed to obtain a +12V power rail and a -12V power rail for several charge signal processing modules.

10. The miniaturized intelligent signal decoupling device suitable for a piezoelectric multi-dimensional force measurement system according to claim 5, characterized in that: It also includes a box, the main controller and the onboard power supply of the intelligent data decoupling unit are integrated on the first PCB circuit board, and the first PCB circuit board is arranged in the middle position of the box; a plurality of charge signal processing modules are respectively integrated on the second PCB circuit board and the third PCB circuit board, the second PCB circuit board and the third PCB circuit board are respectively arranged in the box on both sides of the first PCB circuit board at intervals, and the human-computer interaction module and the communication module are embedded in different surfaces of the box; a plurality of input channels are respectively arranged on the second PCB circuit board and the third PCB circuit board, and the input channels are used to be electrically connected to a plurality of piezoelectric six-dimensional force sensors in a one-to-one correspondence; shielding covers are arranged on both sides of the second PCB circuit board and the third PCB circuit board; adjacent PCB circuit boards are fixed and spaced by copper columns.

Images