Implementation of soft and hard compensation method for embedded pressure monitoring system of oil cylinder pressure

    With the development of modern science and technology, the performance of special vehicles is getting higher and higher, and the system structure is becoming more and more complex and precise. The use of embedded sensors to realize real-time online monitoring of many working parameters is an inevitable trend in the development of vehicle status monitoring and fault diagnosis systems in the future [1]. At the same time, due to the compact structure of the vehicle, it is required to realize a small embedded sensor monitoring system. Otherwise, many fault points cannot be directly monitored and can only be obtained by measuring the relevant parameters of the periphery. The correctness of the conversion result is unknown, which may lead to greater economic losses. With the development of microelectronics technology and signal processing technology, the realization of a small pressure monitoring system has become possible. This paper adopts advanced pressure sensor devices, combined with precision amplifier circuits and low-power high-performance processors to form an embedded special vehicle planetary transmission mechanism operating pressure real-time monitoring system, and through the comprehensive application of measurement circuits and compensation algorithms, high-precision error compensation of the monitoring system is realized.

    1 Implementation of soft and hard compensation method for embedded pressure monitoring system

    1.1 Introduction to pressure monitoring system

    The planetary transmission mechanism of special vehicles is located in the large box of the integrated transmission device. The surrounding space is small and part of it is immersed in lubricating oil with a temperature as high as 135 °C. Through demonstration, a small isolation diaphragm pressure sensor is selected and installed on the oil channel of the cylinder accessory. After the signal is amplified by a precision instrument amplifier, it is quantified by a high-effective bit analog-to-digital converter (ADC), and the data is processed by ATMEL's AVR series microcontroller (MCU) [2]. The processing results are then provided to the operator through the CAN 2.0 bus to achieve the purpose of protecting the vehicle's integrated transmission device. The system principle block diagram is shown in Figure 1.

    Since the operating temperature range of the system is relatively wide (-20 ℃ ~ 135 ℃), temperature changes have the greatest impact on the measurement error of the system. This paper focuses on studying and implementing the compensation method for the measurement error caused by temperature drift [3]. The compensation method of the entire system includes two parts: circuit hard compensation and algorithm soft compensation. The hard compensation includes the process modulation compensation and amplification circuit compensation of the sensor itself; the soft compensation is achieved by embedding the B-spline temperature compensation algorithm in the MCU.

    1.2 Hard compensation circuit design

    The system uses a silicon piezoresistive pressure sensor, which has the characteristics of small size, high sensitivity, and high resolution, and is widely used. However, temperature drift is the biggest weakness of silicon piezoresistive sensors, which includes zero point temperature drift and sensitivity temperature drift. Since the resistance values ​​of the four resistors that make up the bridge cannot be completely consistent, when the input pressure is zero, the bridge output is not zero and has a zero point offset. Sensitivity temperature drift is mainly caused by the change of the piezoresistance coefficient of the semiconductor material with the change of temperature. Generally speaking, the sensitivity of the piezoresistive sensor decreases with the increase of temperature [4].

    This system uses a small silicon piezoresistive sensor from a US company, with a maximum range of 300 psi (1 psi = 6.895 kPa), an output voltage of 0~100 mV, and a nonlinearity of ±0.1%. The sensor achieves temperature compensation and zero deviation adjustment of the sensor by laser trimming the thick film resistor on the ceramic base. The laser trimmed resistor provided inside is used to adjust the gain of the external amplifier to ensure the ±0.1% interchangeability range of the sensor. The circuit schematic is shown in Figure 2.

    Since the minimum resolution of the sensor is at the microvolt level, it is very easy to generate interference during transmission and measurement, resulting in distorted results . Therefore, a high-precision, high common-mode rejection ratio measurement amplifier circuit must be used to amplify small signals.

    This system uses the INA128 instrumentation amplifier with differential input and closed-loop gain unit as the front-end amplifier circuit for hard compensation. The two input terminals of the instrumentation amplifier have balanced impedance and high resistance, low input bias current , and low output impedance. The common-mode rejection is 100 dB, which can reduce any error caused by the common-mode level to 100 dB. Its internal structure is shown in Figure 3, which is exactly the same as the amplifier in the hard compensation circuit shown in Figure 2.

    1.3 Implementation of soft compensation algorithm

    After the system has been hard-compensated, the voltage signal is sampled by a 24-bit quantized analog-to-digital converter (ADC) and sent to the MCU for subsequent processing. The MCU uses ATMEL's ATmega32, which has a data throughput of up to 1 MIPS/MHz . The data processing capability of the MCU is used to implement a temperature drift compensation processing algorithm based on B-spline. The reason why the B-spline curve is used to fit the temperature coefficient of the pressure sensor is that the B-spline curve has a local control characteristic, and the curve only changes shape near the changed control point; in addition, the B-spline curve can add control points at will without increasing the order of the curve. The number of control points can be selected for different applications to meet different fitting requirements [5].

    A spline is a piecewise polynomial function. The expression of a k-order (k-1) B-spline curve is:

    When the denominator is zero, the value of the defined fraction is zero. Where ti represents the control point node value, which controls the shape of the curve, and the node value is from t0 to tn+4. In this system, a non-closed curve is used, so the ti value rule is as follows:

    The compensation algorithm based on B-spline is implemented in MCU. Considering that B-spline is a non-closed curve, the number of control points determined by equations (4) and (5) should be more than 4. Considering the real-time performance of the compensation algorithm, certain requirements are also put forward for the calculation speed, and the control point should not be too large. Based on the above reasons, n=6 is taken, and the sensor data is sampled. The calculation is performed through the single-chip microcomputer . According to the calculation formula of the B-spline curve, the control node value ti is derived, and the solution program of the cubic B-spline harmonic function is written. A compensation coefficient C(u) is calculated for each control point, and the compensation coefficient is written into the corresponding Flash register. When the system is working normally, the temperature value of the pressure monitoring system is obtained through the vehicle temperature sensor, and the compensation coefficient of the corresponding register is read. The measured value is compensated to obtain the corrected pressure output.

    2 Experimental results of soft and hard compensation methods

    During the experiment, the pressure sensor was placed in a 150 psi constant pressure environment, and the output voltage of the pressure sensor signal after amplification was measured. If the system is not affected by temperature drift, the output voltage should be stable at 2.5 V in theory. The measured data after hard compensation at different temperatures are shown in Table 1.

    Draw a curve according to the data in Table 1 and compare it with the ideal output, as shown in Figure 4. It can be clearly seen that the output does not change with temperature in a straight line, but in a nonlinear manner, which means that after the previous hard compensation, although the nonlinearity can be limited to 0.1% of the full scale, it still does not meet the requirements for systems with high precision requirements, and further soft compensation is required.

    The hard-compensated data is then corrected in the MCU using the B-spline soft compensation algorithm. The corrected results are shown in Table 2.

    It can be seen from Table 2 that after the system is soft-compensated, the measured values ​​of the seven temperature points have much smaller errors than the theoretical values. Compared with only hard compensation, the error can be reduced to about 1/5. The error curves after soft and hard compensation are shown in Figure 5.

    This paper focuses on the temperature drift error compensation method for high-precision pressure monitoring systems in new special vehicles. Compared with traditional methods, the main feature of this method is the combination of hard compensation and soft compensation. Hard compensation includes the process modulation compensation and amplification circuit compensation of the sensor itself; soft compensation is to perform temperature compensation by embedding B-spline curve fitting in the MCU. Compared with other methods, this method is easy to implement and has a small fitting error, which greatly improves the accuracy of the system and improves the robustness and stability of the system. The system has been tried in a certain prototype vehicle, with good operating conditions and high compensation accuracy, and has great promotion and application value.