CN-121994340-A - Digital model verification method for reliability of weighing sensor

CN121994340ACN 121994340 ACN121994340 ACN 121994340ACN-121994340-A

Abstract

The invention relates to the technical field of sensor detection, in particular to a method for verifying a digital model of the reliability of a weighing sensor. The technical scheme includes that a digital intelligent modeling module is constructed and comprises a processor, a D/A converter, an A/D converter, an excitation switch circuit and a signal sampling circuit, and a first standard excitation signal is applied to a sensor through the digital intelligent modeling module in a sensor zero load state. The invention realizes on-line and quantitative evaluation of the recessive performance degradation of the sensor on the premise of not disassembling and stopping production by a digital model and reverse verification technology, can discover soft faults which are difficult to identify by a traditional method in an early stage, supports predictive maintenance decisions, prolongs the effective service life of the sensor by dynamic compensation, and improves the fault positioning and operation and maintenance efficiency of a multi-sensor system, thereby obviously improving the comprehensive utilization rate of equipment and reducing the maintenance cost of a full life cycle.

Inventors

LIU YING
ZENG BO
XIAN XIUJUAN
DANG ZHENGQIANG

Assignees

中国测试技术研究院

Dates

Publication Date: 20260508
Application Date: 20260119

Claims (9)

1. The method for verifying the digital model of the reliability of the weighing sensor is characterized by comprising the following steps of: S1, constructing a digital intelligent modeling module, wherein the module comprises a processor, a D/A converter, an A/D converter, an excitation switch circuit and a signal sampling circuit; S2, under the zero load state of the sensor, a first standard excitation signal is applied to the sensor through a digital intelligent modeling module, and is acquired and processed through a signal sampling circuit and an A/D converter to obtain first reverse response data, and a first reverse curve difference value VC1 is obtained based on the first reverse response data; s3, when the load of the sensor is larger than a preset threshold value, a second standard excitation signal is applied to the sensor through a digital intelligent modeling module, and is acquired and processed through a signal sampling circuit and an A/D converter to obtain second reverse response data, and a second reverse curve difference value VC2 is obtained based on the second reverse response data; S4, storing the first reverse curve difference value VC1 as a reference value in a memory of a digital intelligent modeling module; s5, repeating the steps S2-S3 in the subsequent reliability verification process to obtain a new reverse curve difference VCn; S6, comparing the VCn with a stored reference value VC1, if the ratio of the absolute value of the difference value of the VCn and the reference value VC1 is within a preset tolerance range, judging that the sensor is normal, and if the absolute value of the difference value is beyond the tolerance range, judging that the sensor is abnormal.
2. The method of claim 1, wherein the digital intelligent modeling module is a stand-alone hardware calibration device or an integrated functional module embedded within a weighing instrument or sensor.
3. The method for verifying a digitized model of the reliability of a load cell of claim 2, wherein the applying of the standard excitation signal in steps S2 and S3 comprises: the processor generates a digital control signal according to a preset internal code value, the digital control signal is converted into an analog current signal through the D/A converter, and the analog current signal is connected into the signal sampling circuit by the excitation switch circuit and is superposed into an output signal loop of the sensor.
4. A method of verifying a digitized sensor reliability model according to claim 3 wherein the predetermined internal code value corresponds to a partial range value of the sensor and the predetermined threshold state is any load point between 30% and 100% of the maximum sensor weight.
5. The method for verifying the digitized model of the reliability of a weighing sensor according to claim 1, wherein the calculation of the inverse curve difference value VC is obtained by performing an integral operation in a range from a zero voltage value V0 to a target load point voltage value Vt based on an output response curve of the sensor under the action of an excitation signal.
6. The method for verifying a digital model of reliability of a load cell according to claim 5, wherein the mathematical model of comparison and determination in step S6 is: | (VC1 - VCn) / VC1 | × 100% ≤ δ% wherein δ% is a preset tolerance percentage threshold.
7. The method of claim 1, wherein when the object to be verified is a weighing system comprising a plurality of sensors, the method further comprises: each sensor is distributed with an independent identifier and is connected with the digital intelligent modeling module through a wireless or wired communication module; the digital intelligent modeling module concurrently applies the standard excitation signals to each sensor, and receives and processes the reverse response data of each sensor; respectively calculating and storing a reference reverse curve difference value of each sensor; in the subsequent verification, the difference value of the new and old reverse curves of each sensor is respectively compared, and the abnormal sensor and the identification thereof are positioned and displayed.
8. The method of claim 1, further comprising the step of calibrating and compensating, wherein the digital intelligent modeling module dynamically compensates the measured output value of the sensor by using current reverse modeling data when the sensor is determined to be abnormal but not completely invalid, so as to maintain the overall measurement accuracy of the system.
9. The method of claim 1, wherein the predetermined tolerance range is dynamically adjusted based on the sensor range, the number of divisions, historical performance degradation data, and the reliability requirements of the application scenario.

Description

Digital model verification method for reliability of weighing sensor Technical Field The invention relates to the technical field of sensor detection, in particular to a method for verifying a digital model of the reliability of a weighing sensor. Background The weighing sensor is used as a core measuring element in the fields of industrial weighing, process control and safety monitoring, and the long-term operation reliability of the weighing sensor is directly related to the production quality, the fairness of trade settlement and the operation safety of equipment. The performance of the sensor can degrade with time, environmental stress (such as temperature, humidity, mechanical vibration and electromagnetic interference) and long-term load, and the sensor is represented by hidden faults such as zero drift, sensitivity change, nonlinear error increase and the like. How to timely and accurately evaluate the health state of the sensor and predict the residual life of the sensor on the premise of not interrupting production and not dismantling equipment is a key technical problem facing industrial field maintenance. Currently, the field reliability judgment and fault detection of a weighing sensor mainly depend on the following traditional methods: And the output signal judging method is to use a universal meter to measure the output millivolt signal of the sensor bridge circuit when the system is in idle load. Whether the sensor has obvious faults or not is roughly judged by comparing whether the output value of the sensor of the same type is in an expected range or stable. This approach relies heavily on experience and cannot identify gradual degradation of performance. And the impedance judging method is to measure the input/output impedance and insulation resistance of the sensor after power failure. If the impedance value is abnormally lowered, it can be judged that the sensor has a 'hard fault' such as short circuit, damp or internal damage. This approach is also insensitive to "soft faults" such as degraded sensitivity, intermittent faults, etc. And the cross substitution and elimination method is to observe the change of the weighing display instrument by connecting the signal simulators or disconnecting the sensor connection lines one by one so as to isolate the fault sensor. The method is complex in operation, the system operation is required to be interrupted, and when the inherent discreteness exists in the sensors in the same batch, the sensors which are at the performance edge and still available are easily misjudged as fault parts, so that unnecessary replacement and cost waste are caused. In combination, the above-described conventional methods have three general inherent drawbacks: the standard is not uniform, and judgment depends on experience of operators or transverse comparison of sensors in the same batch, and objective and quantitative uniform performance standard is lacked. The scene is difficult to reproduce, and the laboratory high-precision calibration is separated from complex comprehensive working conditions such as temperature and humidity, vibration, electromagnetic interference and the like on site, so that the measurement accuracy of the laboratory is not equal to the stability of the field. And the hidden fault missed judgment is that only sudden hard faults such as open circuit, short circuit and the like can be detected, and the slow degradation, slight drift and intermittent abnormality of the sensor performance are basically disabled. In the software level, although the common threshold alarming method can set the upper limit and the lower limit, the fixed threshold is very easy to generate false alarm or missing alarm in the face of the nonlinear aging curve of the sensor. The method has the root problems that the traditional method simplifies the complex failure mechanism of the sensor into static single-point signal comparison, and the lack of continuous and quantitative tracking and modeling analysis on the full life cycle of the sensor 'performance degradation', directly causes the passive swing of equipment maintenance strategies between excessive maintenance and sudden shutdown, and ensures that the sensor 'with disease on duty' becomes normal until quality or safety accidents are caused to be traced back afterwards. Therefore, the application provides a method for verifying the reliability digital model of the weighing sensor. Disclosure of Invention Aiming at the problem that the traditional method simplifies the complex failure mechanism of the sensor into static single-point signal comparison in the background technology, the continuous, quantitative tracking and modeling analysis on the full life cycle of the performance degradation of the sensor are lacked, and a digital model verification method for the reliability of the weighing sensor is provided. The technical scheme of the invention is that the method for verifying the digital model of the reliability