Bus type real-time multi-channel weighing detection system and method

Abstract

The invention discloses a bus type real-time multi-channel weighing detection system and method, and relates to the technical field of industrial control, and the method comprises the steps: arranging a multi-point non-equidistant configuration formed by a head sampling point P1, a middle sampling point P2 and a tail sampling point P3 in each weighing channel of a bus type multi-channel weighing instrument; in combination with bus timestamp synchronization and a priority interrupt response strategy of a real-time operating system RTOS, original weighing curve data W can be efficiently acquired in unit time, and asynchronous data packaging is performed through a bus type data path and transmitted to a central industrial control processing module. Through introduction of the sampling point arrangement mode and the communication scheduling mechanism, the original weighing data sequence has time continuity and channel alignment consistency, a basic guarantee is provided for extraction and dynamic evaluation of a subsequent state data set, and the data processing capacity and response speed of the system under the conditions of high frequency and high channel number are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of industrial control technology, specifically to a bus-based real-time multi-channel weighing and detection system and method. Background Technology

[0002] With the continuous development of industrial control automation and intelligent manufacturing technologies, especially in fields such as pharmaceuticals, chemicals, and food where high precision in fluid proportioning is required, multi-channel synchronous liquid filling and precision weighing have gradually become core technologies. To achieve parallel control and data acquisition of multiple liquid channels, multi-channel weighing systems with a bus architecture have been widely adopted in related applications. These systems complete data transmission and control scheduling across multiple channels through a unified communication bus, featuring high channel density, strong real-time performance, and flexible deployment, making them a crucial infrastructure for achieving high-precision, continuous weighing and detection. Supported by industrial control systems, developing a bus-based real-time multi-channel weighing and detection method with self-sensing capabilities and dynamic sensitivity correction is of significant application value for further improving the robustness, adaptability, and accuracy stability of multi-channel systems.

[0003] Currently, in multi-channel liquid weighing and detection systems, static calibration is commonly used to set the weighing sensitivity parameters of each channel. However, due to the coupled effects of liquid flow conditions, pipeline resistance, sensor response differences, and changes in liquid density, the originally set static sensitivity parameters cannot continuously adapt to fluctuations during operation, leading to sensitivity drift and inconsistencies in accuracy across different channels during weighing. Furthermore, existing systems typically lack real-time response and compensation mechanisms for minute changes in flow rate and nonlinear disturbances, which can easily result in decreased accuracy, increased fluctuations, and data distortion under low flow rate or pulsating flow conditions.

[0004] The root cause of these problems lies in the fact that existing weighing systems cannot capture the changing trends of liquid states over time in real time, particularly lacking the dynamic sensing capability for density disturbances, sudden changes in flow rate, and local micro-vibrations. Because the system lacks a feedback model for quantitatively assessing changes in channel states and adaptively adjusting sensitivity, it cannot effectively and dynamically calibrate sensitivity parameters should a channel experience local sensor aging, pipeline bubble interference, or sudden changes in drug characteristics. This leads to systematic errors in the weighing results. Furthermore, data mismatch between channels will further affect batch consistency and may even lead to serious abnormal consequences such as unbalanced drug ratios, compliance risks, or failure of dosage control. This is especially problematic in high-precision pharmaceutical filling or chemical reagent preparation scenarios, easily causing quality fluctuations in downstream processes or product scrapping. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a bus-based real-time multi-channel weighing and detection system and method, which solves the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution, comprising the following steps: S1. By setting several sampling points in each weighing channel of the bus-type multi-channel weighing instrument, the original weighing data sequence of each channel is acquired in real time, and the original weighing data sequence is input to the industrial control processing module for preprocessing and extraction of the status dataset. S2. Based on the state dataset, calculate the state drift index Py, and set the drift threshold Pth to perform a preliminary comparison and evaluation with the state drift index Py. If the preliminary comparison and evaluation indicates abnormal fluctuation, trigger the dynamic sensitivity prediction mechanism to calculate and output the sensitivity offset index △K. S3. Set the offset threshold Kth and the sensitivity offset index △K for a second comparison evaluation. If the sensitivity deviates abnormally in the second comparison evaluation, the redundant channel compensation mechanism is triggered.

[0007] Preferably, S1 includes S11; S11. Three sampling points located at different structural positions are set on each weighing channel, namely head sampling point P1, middle sampling point P2 and tail sampling point P3. The three sampling points are distributed in a multi-point non-equidistant configuration on the weighing channel structure. Each sampling point collects force information through the installed weighing sensor, and the analog signal acquisition module connected to the weighing sensor performs analog-to-digital conversion processing to obtain the original weighing curve data W per unit time. The acquisition operation of each sampling point is synchronized with the bus timestamp and executed in conjunction with the priority interrupt response strategy of the real-time operating system (RTOS). The original weighing curve data W of each channel per unit time is sorted according to the timestamp order and the original weighing data sequence is obtained by summarizing.

[0008] Preferably, S1 further includes S12; S12. The original weighing data sequence collected from the collection point is packaged according to the bus data frame format and sent to the central industrial control processing module through the communication bus connected to the weighing channel in an asynchronous transmission protocol. The communication bus is a bus-type data path with multi-channel real-time transmission capability, supporting data identification with timestamps and frame verification mechanism. At the same time, a buffer management and task scheduling mechanism is adopted during the transmission process.

[0009] Preferably, S1 further includes S13; S13. Preprocess the original weighing data sequence in the central industrial control processing module and extract the status dataset. The state dataset includes first-order derivative curve values. And the liquid density parameter p of the i-th channel i ; The preprocessing involves performing denoising and smoothing using a multi-point sliding window method based on the original weighing curve data W of each channel per unit time in the original weighing data sequence. Then, based on the denoised and smoothed original weighing curve data W of each channel per unit time, first-order difference calculations are performed to obtain the first-order derivative curve values. Where d represents the integral function, W i This represents the original weighing curve data of the i-th channel; simultaneously, based on the original weighing curve data W and the liquid volume, a density inversion calculation process is performed to extract the liquid density parameter p of the liquid in each channel.

[0010] Preferably, S2 includes S21; S21. The first-order derivative curve value of each channel extracted from the state dataset. Extract the first derivative curve value within the set sampling window duration. The maximum and minimum values ​​in the sample are used to obtain the state drift index Py, which represents the degree of change in channel weighing, by dividing the difference between the maximum and minimum values ​​by the sampling window duration Δt.

[0011] Preferably, S2 further includes S22; S22. Based on historical multi-channel weighing data, the drift index Py of multiple channels in both stable and abnormal states is collected, and a drift threshold Pth is set. The drift threshold Pth is set by the upper limit of the stable state drift index Py. A preliminary comparison and evaluation is performed between the real-time acquired state drift index Py and the drift threshold Pth to determine the stability of the current channel weighing change. Based on the preliminary comparison and evaluation results, a dynamic sensitivity prediction mechanism is triggered. The specific evaluation content is as follows: When the state drift index Py ≥ drift threshold Pth, it indicates that the current channel weighing is unstable, and the dynamic sensitivity prediction mechanism is triggered at this time. When the state drift index Py < drift threshold Pth, it indicates that the channel weighing change is stable, and the current static and dynamic sensitivity correction process is adopted at this time.

[0012] Preferably, S2 further includes S23; S23. After triggering the dynamic sensitivity prediction mechanism, the liquid density parameter p of the i-th channel extracted from the industrial control processing module is used. i The sensitivity offset index ΔK is obtained by correlating the state drift index Py with the current channel sensitivity to measure the fluctuation difference between the current channel sensitivity and the neighboring channels.

[0013] Preferably, S3 includes S31; S31. Based on the upper limit of the acceptable range of channel sensitivity deviation, set an offset threshold Kth, then compare and evaluate the real-time acquired sensitivity deviation index ΔK with the offset threshold Kth to determine the current sensitivity deviation of the weighing channel, and trigger the redundant channel compensation mechanism based on the results of the second comparison and evaluation; the specific evaluation content is as follows: When the sensitivity offset index ΔK > offset threshold Kth, it indicates an abnormal deviation, and the redundant channel compensation mechanism is triggered to correct the current channel weighing result. When the sensitivity offset index ΔK ≤ offset threshold Kth, it indicates that the deviation is normal, and the current channel weighing result should be maintained.

[0014] Preferably, S3 further includes S31; S31. After triggering the redundant channel compensation mechanism, the original weighing value W of the i-th channel is used... i (t) is linearly fused with the weighted average weighing value Wneigh of the adjacent channel to obtain the compensated and corrected weighing value W'.

[0015] A bus-based real-time multi-channel weighing detection system includes a weighing sampling module, a weighing sensitivity analysis module, and a compensation analysis module; The weighing sampling module sets several sampling points in each weighing channel of the bus-type multi-channel weighing instrument to acquire the original weighing data sequence of each channel in real time, and inputs the original weighing data sequence to the industrial control processing module for extraction and preprocessing to extract the state dataset. The weighing sensitivity analysis module calculates the state drift index Py based on the state dataset, and sets a drift threshold Pth to perform a preliminary comparison and evaluation with the state drift index Py. If the preliminary comparison and evaluation indicates abnormal fluctuation, the dynamic sensitivity prediction mechanism is triggered to calculate and output the sensitivity offset index △K. The compensation analysis module performs a secondary comparison evaluation by setting an offset threshold Kth and a sensitivity offset index ΔK. If the sensitivity of the secondary comparison evaluation deviates abnormally, the redundant channel compensation mechanism is triggered.

[0016] This invention provides a bus-based real-time multi-channel weighing and detection system and method. It has the following beneficial effects: (1) This method, by setting up a multi-point non-equidistant configuration consisting of a head sampling point P1, a middle sampling point P2, and a tail sampling point P3 in each weighing channel of a bus-type multi-channel weighing instrument, combined with bus timestamp synchronization and the priority interrupt response strategy of the real-time operating system (RTOS), can efficiently collect the original weighing curve data W per unit time and transmit it asynchronously to the central industrial control processing module through the bus-type data path. The introduction of the above sampling point layout and communication scheduling mechanism enables the original weighing data sequence to have temporal continuity and channel alignment consistency, providing a basic guarantee for the subsequent extraction and dynamic evaluation of the state dataset, and significantly improving the system's data processing capability and response speed under high frequency and high channel number conditions.

[0017] (2) This method constructs a state dataset, extracts data containing the first-order derivative curve value and the liquid density parameter pi, calculates the state drift index Py based on the range of the first-order derivative curve value and the sampling window duration Δt, and sets a drift threshold Pth by combining the upper limit of the state drift index under the stable state of the historical channel. This allows for a preliminary comparative evaluation of real-time weighing fluctuations. When an abnormal drift occurs, a sensitivity correction factor is further calculated based on the density change rate and the state disturbance index, and a sensitivity offset index ΔK is calculated by comparing it with the average correction factor of neighboring channels. After a secondary comparative evaluation with the offset threshold Kth, the method accurately identifies whether the sensitivity deviation exceeds the acceptable range. This two-stage dynamic evaluation mechanism has the ability to distinguish between transient channel fluctuations and systematic deviations, thereby effectively improving the robustness and accuracy of sensitivity prediction.

[0018] (3) When the sensitivity offset index ΔK exceeds the offset threshold Kth, the system triggers the redundant channel compensation mechanism and adaptively sets the compensation ratio factor based on the degree of sensitivity offset. A linear fusion is performed between the original weighing value Wi(t) of the i-th channel and the weighted average weighing value W̄neigh(i) of the neighboring channels to calculate the compensated weighing value W′i(t). Where 0.1 < ΔKi < 0.2, The value increases linearly to 0.5; when ΔKi ≥ 0.2, A step-wise strategy was adopted, with the value set between 0.5 and 0.8, to achieve differentiated control of the compensation intensity. This compensation strategy not only ensures that the unstable channel can still output continuous and reliable corrected data during data anomalies, but also improves the overall weighing system's self-recovery capability for single-point failures and the redundancy and fault tolerance performance of multi-channel data. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the steps of a bus-based real-time multi-channel weighing and detection method according to the present invention.

[0020] Figure 2 This is a schematic diagram of the bus-type real-time multi-channel weighing and detection system of the present invention.

[0021] Figure 3 This is a distribution chart of the sensitivity offset index ΔK. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] Example 1 Please see Figure 1 and Figure 3 This invention provides a bus-based real-time multi-channel weighing detection method. To achieve the above objectives, this invention employs the following technical solution, comprising the following steps: S1. By setting several sampling points in each weighing channel of the bus-type multi-channel weighing instrument, the original weighing data sequence of each channel is acquired in real time, and the original weighing data sequence is input to the industrial control processing module for preprocessing and extraction of the status dataset. S2. Based on the state dataset, calculate the state drift index Py, and set the drift threshold Pth to perform a preliminary comparison and evaluation with the state drift index Py. If the preliminary comparison and evaluation indicates abnormal fluctuation, trigger the dynamic sensitivity prediction mechanism to calculate and output the sensitivity offset index △K. S3. Set the offset threshold Kth and the sensitivity offset index △K for a second comparison evaluation. If the sensitivity deviates abnormally in the second comparison evaluation, the redundant channel compensation mechanism is triggered.

[0024] In this embodiment, the method addresses the problems of weighing fluctuations, inconsistent channel responses, and error accumulation that easily occur in multi-channel weighing systems under dynamic scenarios. It proposes a systematic solution process from data acquisition and configuration to data processing and anomaly compensation. First, three non-equidistant sampling points are set on each weighing channel structure to collect the original weighing curve data W from the head, middle, and tail sections, respectively, avoiding single-point distortion caused by different liquid flow states at different locations. Then, the sampled data is transmitted to the central industrial control processing module in real time via asynchronous bus transmission, and a timestamp and frame verification mechanism is used to ensure data alignment and integrity. In the control module, the first-order derivative curve value is extracted using the first-order derivative method and density inversion algorithm. A state dataset is constructed using the liquid density parameter p, and a state drift index Py is calculated to assess the degree of fluctuation in the weighing state in real time. When the state drift index Py exceeds the set drift threshold Pth, it indicates that the channel weighing curve is changing drastically, triggering a sensitivity prediction mechanism. Combining the current density change and fluctuation trend, a sensitivity offset index ΔK is calculated to determine whether there is an abnormal deviation between the channel response and adjacent channels. If the sensitivity offset index ΔK further exceeds the offset threshold Kth, it indicates that the channel may have a risk of measurement inaccuracy, thus triggering a redundant channel compensation mechanism. By fusing the weighing values ​​of the current channel and adjacent channels, the compensation ratio factor γ is dynamically adjusted to generate a compensated and corrected weighing value W'. This method, through hierarchical detection, quantitative evaluation, and intelligent compensation mechanisms, significantly improves the robustness and accuracy of the system under complex actual working conditions such as disturbances, noise, and imbalances, ensuring the continuity, reliability, and engineering applicability of weighing data. Example 2 Please see Figure 1 Specifically: S1 includes S11; S11. Three sampling points located at different structural positions are set on each weighing channel, namely head sampling point P1, middle sampling point P2 and tail sampling point P3. The three sampling points are distributed in a multi-point non-equidistant configuration on the weighing channel structure. Each sampling point collects force information through the installed weighing sensor, and the analog signal acquisition module connected to the weighing sensor performs analog-to-digital conversion processing to obtain the original weighing curve data W per unit time. The acquisition operation of each sampling point is synchronized with the bus timestamp and executed in conjunction with the priority interrupt response strategy of the real-time operating system (RTOS) to achieve non-blocking scheduling during data transmission. The original weighing curve data W of each channel per unit time is sorted according to the timestamp order and the original weighing data sequence is obtained by summarizing.

[0025] S1 also includes S12; S12. The raw weighing data sequence collected from the sampling points is packaged according to the bus data frame format and sent to the central industrial control processing module via the communication bus connected to the weighing channel using an asynchronous transmission protocol. The communication bus is a bus-type data path with multi-channel real-time transmission capability, supporting data identification with timestamps and frame verification mechanisms to ensure frame synchronization, channel alignment, and real-time arrival of data from each channel. At the same time, a buffer management and task scheduling mechanism is adopted during the transmission process to ensure that the transmission delay remains within a controllable range under high sampling rate conditions.

[0026] S1 also includes S13; S13. Preprocess the original weighing data sequence in the central industrial control processing module and extract the status dataset. The state dataset includes first derivative curve values. And the liquid density parameter p of the i-th channel i ; Preprocessing involves denoising and smoothing the original weighing curve data W for each channel per unit time in the original weighing data sequence using a multi-point sliding window method. Then, based on the denoised and smoothed original weighing curve data W for each channel per unit time, first-order difference calculations are performed to obtain the first-order derivative curve values. Where d represents the integral function, W i The original weighing curve data of the i-th channel is used to obtain an estimate representing the flow rate trend of the channel; at the same time, based on the original weighing curve data W and the liquid volume, a density inversion calculation process is performed to extract the liquid density parameter p of the liquid in each channel.

[0027] In this embodiment, the method effectively avoids weighing signal deviation caused by velocity gradients or local disturbances in different areas of the pipe by setting three non-equidistant sampling points at different locations in each weighing channel, thus enhancing the comprehensiveness and stability of the overall weighing response of the channel. Furthermore, in conjunction with the timestamp synchronization and interrupt scheduling mechanism under the RTOS system, non-blocking data transmission under high sampling rate conditions can be achieved, effectively reducing the risk of frame loss due to thread waiting. In step S12, the original weighing data sequence is packaged according to the bus frame format and asynchronously transmitted through a communication bus with multi-channel concurrency capability. This not only improves the system's data throughput but also ensures channel data alignment accuracy through the frame verification mechanism, making it particularly suitable for highly dynamic environments with a large number of channels and dense data flow. In step S13, denoising and smoothing are performed using the sliding window method, and the first derivative curve value and liquid density parameter pi are calculated. This more accurately characterizes the flow trend and actual mass state of the liquid in each channel, providing high-confidence input data for subsequent state assessment and sensitivity prediction. The above-mentioned settings enable the weighing system to have higher anti-interference capabilities, data accuracy, and engineering adaptability, effectively supporting complex dynamic weighing needs under multi-channel structures.

[0028] Example 3 Please see Figure 1 and Figure 3 Specifically: S2 includes S21; S21. The first-order derivative curve value of each channel extracted from the state dataset. Extract the first derivative curve value within the set sampling window duration. The maximum and minimum values ​​in the sample are obtained, and the state drift index Py, which represents the degree of change in channel weighing, is obtained by dividing the difference between the maximum and minimum values ​​by the sampling window duration Δt. The state drift index Py is calculated and output using the following algorithm formula; ; In the formula, Py i Let represent the state drift exponent of the i-th channel, Δt represent the sampling window duration, max represent the maximum value function, and min represent the minimum value function; The state drift index Py is designed to describe the consistency and volatility of the weighing change curve of a certain channel in a multi-channel weighing system within a given time window. This index draws on the range analysis method in statistics to achieve a normalized quantitative assessment of the unsteady disturbances inside the weighing system. This dimension represents the rate of change of mass acceleration, which can be physically interpreted as "weighted disturbance acceleration," and is reasonable for characterizing the intensity of fluctuations.

[0029] S2 also includes S22; S22. Based on historical multi-channel weighing data, the drift index Py of multiple channels in both stable and abnormal states is collected. A drift threshold Pth is set, which is based on the upper limit of the stable state drift index Py. A preliminary comparison and evaluation is performed between the real-time acquired drift index Py and the drift threshold Pth to determine the stability of the current channel's weighing changes. Based on the preliminary comparison and evaluation results, a dynamic sensitivity prediction mechanism is triggered. The specific evaluation content is as follows: When the state drift index Py ≥ drift threshold Pth, it indicates that the current channel weighing is unstable, and the dynamic sensitivity prediction mechanism is triggered at this time. When the state drift index Py < drift threshold Pth, it indicates that the channel weighing change is stable, and the current static and dynamic sensitivity correction process is adopted at this time.

[0030] S2 also includes S23; S23. After triggering the dynamic sensitivity prediction mechanism, the liquid density parameter p of the i-th channel extracted from the industrial control processing module is used. i The sensitivity offset index ΔK is obtained by correlating the state drift index Py with the current channel sensitivity to measure the fluctuation difference between the current channel sensitivity and the neighboring channels. The sensitivity offset index ΔK is calculated and output using the following algorithm formula; ; In the formula, △K i Kneigh represents the sensitivity offset index of the i-th channel. idenoted as the average value of the sensitivity correction factor of the i-th channel's neighboring channels, calculated by averaging multiple spatially adjacent channels. 'a' represents the density change sensitivity gain coefficient, the direct influence of density change on the sensitivity correction factor, and is a type of proportional amplification factor. Its actual value needs to be calibrated through system experiment fitting, and its range is between 0.5 and 3.0. It is commonly used to adjust the relative weights of various inputs in the prediction function. 'b' represents the state perturbation coupling factor, used to weight the state drift index Py on the sensitivity correction, acting as a "state suppression / excitation" factor. That is, when channel perturbations are severe, the correction amplitude is appropriately increased or decreased; its value is a positive real number less than 1 to avoid over-correction and maintain system stability. The sensitivity correction factor is designed based on the dynamic adaptive correction model and nonlinear predictive correction model in control engineering, and is improved by combining the concept of "state disturbance term" in physical processes. The formula has the structure of "baseline sensitivity + state-driven offset" to ensure that both the current state and the fluctuation trend are taken into account. It is an improved nonlinear predictive adjustment function. Furthermore, it is compared with the consistency of neighboring channels to identify potential anomalies and trigger the next compensation strategy. p i The unit is kg / m³ 3 , The unit is kg / (m³). 3 ·s); Py: Unit is kg / s 2 ; a, b: each has a unit m 3 / kg and s 2 / kg, to ensure that all product terms are dimensionless; The overall result shows that the adjacent channels have dimensionless amplification factors, which satisfies the output parameter type, conforms to the principle of physical dimension conservation, and is reasonable and feasible for engineering.

[0031] In this embodiment, the method sets a fixed sampling window duration in S21, extracts the maximum and minimum values ​​of the first-order derivative curve for each channel, and calculates the state drift index Py. This design is based on stability quantification modeling of the dynamic change characteristics of the weighing curve, and is particularly suitable for multi-channel systems where liquid flow exhibits non-steady-state fluctuations. Without this range normalization assessment, it would be difficult to accurately determine whether the weighing channel exhibits abnormal fluctuations. For example, when the liquid causes short-term oscillations in the weighing data due to disturbances such as bubbles, impacts, or local temperature differences, without introducing the state drift index Py, it would be impossible to distinguish the essential difference between such disturbances and normal slow changes. Therefore, the state drift index Py reflects the "second-order derivative trend of mass change" in a physical sense, i.e., the "acceleration of weighing disturbance," and can more realistically characterize the intensity of data fluctuations. In S22, through statistical modeling of the channel state drift index Py in historical sampling, stable and abnormal states are distinguished, and a reasonable drift threshold Pth is further set, enabling the system to adaptively determine the stability of the current channel. This method of dynamically calibrating thresholds based on historical data is more adaptable and fault-tolerant than the traditional fixed threshold method. For example, in high-frequency disturbance industrial environments, using a fixed static threshold may lead to frequent misjudgments, resulting in erroneous triggering of the compensation process and a decline in system performance. By using an adaptive setting based on the upper limit of the steady-state Py, the activation threshold of the "dynamic sensitivity prediction mechanism" can be precisely controlled, ensuring that the subsequent compensation process is triggered only in cases of genuine abnormal fluctuations, thus improving the accuracy and stability of the system response. In step S23, after the dynamic sensitivity prediction mechanism is triggered, the liquid density change rate and the intensity of the state disturbance are introduced as two major variables to construct a sensitivity correction function and output a sensitivity offset index ΔK. This offset index not only reflects the degree of state disturbance in the current channel but also integrates the sensitivity consistency deviation with neighboring channels, possessing a strong self-correction capability for multi-channel systems. For example, when a channel produces an abnormal weighing curve due to local mechanical loosening, pipe wall deposits, or structural inconsistencies, this channel will show a significant deviation in sensitivity response from surrounding channels. ΔK can identify such abnormal signals immediately, effectively preventing them from interfering with the overall weighing results. Overall, the above implementation process has the following significant effects: First, the introduction of the state drift index Py improves the quantitative identification capability of weighing curve disturbances; second, the use of a dynamic drift threshold Pth based on historical data enhances the system's adaptability to complex working conditions; and third, the sensitivity offset index ΔK enables multi-channel data consistency monitoring and anomaly identification. The collaborative design of these sub-modules enables the entire system to maintain high-precision weighing output even in highly dynamic and strongly disturbed environments, significantly improving the robustness, stability, and intelligent decision-making capabilities of the weighing system.

[0032] Example 4 Please see Figure 1Specifically: S3 includes S31; S31. Based on the upper limit of the acceptable range of channel sensitivity deviation, set an offset threshold Kth, then compare and evaluate the real-time acquired sensitivity deviation index ΔK with the offset threshold Kth to determine the current sensitivity deviation of the weighing channel, and trigger the redundant channel compensation mechanism based on the results of the second comparison and evaluation; the specific evaluation content is as follows: When the sensitivity offset index ΔK > offset threshold Kth, it indicates an abnormal deviation, and the redundant channel compensation mechanism is triggered to correct the current channel weighing result. When the sensitivity offset index ΔK ≤ offset threshold Kth, it indicates that the deviation is normal, and the current channel weighing result should be maintained.

[0033] S3 also includes S31; S31. After triggering the redundant channel compensation mechanism, the original weighing value W of the i-th channel is used... i (t) is linearly fused with the weighted average weighing value Wneigh of the adjacent channel to obtain the compensated and corrected weighing value W'; The compensated and corrected weighing value W' is calculated and output using the following formula; ; In the formula, W' i (t) represents the compensated and corrected weighing value of the i-th channel at time t. Wneigh represents the compensation ratio factor. i This represents the weighted average weight of the near channels in channel i; Wherein: compensation ratio factor Based on the sensitivity offset index ΔK, the following condition is met: When 0.1 < ΔK i When <0.2, the compensation ratio factor Increment from 0.3 to 0.5 using a linear strategy; When ΔK i When ≥0.2, the compensation ratio factor The step strategy is set between 0.5 and 0.8; Compensation ratio factor The larger the value, the more severe the distortion of the current channel's weighing signal, and the higher the redundancy weight allocation to adjacent channels, in order to ensure the stability and accuracy of continuous data output.

[0034] In this embodiment, the method effectively identifies whether the current channel's weighing sensitivity has exceeded the system's allowable range by comparing the offset threshold Kth set in S31 with the sensitivity offset index ΔK. This secondary judgment mechanism is designed to address the problem of misjudgment caused by instantaneous fluctuations or interference signals, avoiding triggering compensation based solely on a single state assessment, thereby improving the robustness of the system's judgment. For example, during fluid weighing, bubbles or liquid surges may cause temporary sensitivity shifts. Without a secondary comparison, this can easily lead to false triggering of the compensation mechanism, introducing unnecessary correction errors. The redundant channel compensation mechanism proposed in S32 compares the original weighing value Wi(t) of the current channel with the weighted average weighing value Wneigh of neighboring channels. i Linear fusion is performed, introducing "local redundancy" from adjacent dimensions of the signal space to correct abnormal channels. Physically, this mechanism is equivalent to constructing a "signal bypass." When the main channel data deviation is significant, neighboring channels intervene with dynamic weights to participate in the joint output of the weighing result. For example, when ΔKi reaches 0.25, the compensation scaling factor γ will be set above 0.6, meaning that more than 60% of the weighing value is provided by neighboring channels, thereby significantly reducing the impact of distorted channels on the total output. The design of the compensation scaling factor adopts a combination of piecewise linear and step scaling strategies to adapt to different levels of anomaly and achieve more flexible control of the compensation intensity. If the compensation factor design is too rigid, it is easy to cause over-correction even for minor anomalies, which will affect the stability of the system. However, by setting a step increase γ when ΔKi≥0.2, severe distortion events can be quickly responded to, and a stable output transition can be achieved. In summary, this implementation process effectively enhances the system's adaptive perception and dynamic correction capabilities for abnormal channel fluctuations, achieving dual assurance of the stability and accuracy of weighing results. It is particularly suitable for industrial-grade weighing systems in scenarios involving multi-channel parallel operation, high-frequency sampling, and fluctuation sensitivity.

[0035] Example 5 Please see Figure 1 and Figure 2 A bus-type real-time multi-channel weighing detection system includes a weighing sampling module, a weighing sensitivity analysis module, and a compensation analysis module; The weighing sampling module sets several sampling points in each weighing channel of the bus-type multi-channel weighing instrument to acquire the original weighing data sequence of each channel in real time, and inputs the original weighing data sequence to the industrial control processing module for preprocessing and extraction of the state dataset. The weighing sensitivity analysis module calculates the state drift index Py based on the state dataset, and sets a drift threshold Pth to perform a preliminary comparison and evaluation with the state drift index Py. If the preliminary comparison and evaluation indicates abnormal fluctuation, the dynamic sensitivity prediction mechanism is triggered to calculate and output the sensitivity offset index △K. The compensation analysis module performs a secondary comparison evaluation by setting an offset threshold Kth and a sensitivity offset index △K. If the sensitivity of the secondary comparison evaluation deviates abnormally, the redundant channel compensation mechanism is triggered.

[0036] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention.

Claims

1. A bus-based real-time multi-channel weighing and detection method, characterized in that: Includes the following steps: S1. By setting several sampling points in each weighing channel of the bus-type multi-channel weighing instrument, the original weighing data sequence of each channel is acquired in real time, and the original weighing data sequence is input to the industrial control processing module for preprocessing and extraction of the status dataset. S2. Based on the state dataset, calculate the state drift index Py, and set the drift threshold Pth to perform a preliminary comparison and evaluation with the state drift index Py. If the preliminary comparison and evaluation indicates abnormal fluctuation, trigger the dynamic sensitivity prediction mechanism to calculate and output the sensitivity offset index △K. S3. Set the offset threshold Kth and the sensitivity offset index △K for a second comparison evaluation. If the sensitivity deviates abnormally in the second comparison evaluation, the redundant channel compensation mechanism is triggered.

2. The bus-type real-time multi-channel weighing and detection method according to claim 1, characterized in that: S1 includes S11; S11. Three sampling points located at different structural positions are set on each weighing channel, namely head sampling point P1, middle sampling point P2 and tail sampling point P3. The three sampling points are distributed in a multi-point non-equidistant configuration on the weighing channel structure. Each sampling point collects force information through the installed weighing sensor, and the analog signal acquisition module connected to the weighing sensor performs analog-to-digital conversion processing to obtain the original weighing curve data W per unit time. The acquisition operation of each sampling point is synchronized with the bus timestamp and executed in conjunction with the priority interrupt response strategy of the real-time operating system (RTOS). The original weighing curve data W of each channel per unit time is sorted according to the timestamp order and the original weighing data sequence is obtained by summarizing.

3. The bus-type real-time multi-channel weighing detection method according to claim 2, characterized in that: S1 further includes S12; S12. The original weighing data sequence collected from the collection point is packaged according to the bus data frame format and sent to the central industrial control processing module through the communication bus connected to the weighing channel in an asynchronous transmission protocol. The communication bus is a bus-type data path with multi-channel real-time transmission capability, supporting data identification with timestamps and frame verification mechanism. At the same time, a buffer management and task scheduling mechanism is adopted during the transmission process.

4. The bus-type real-time multi-channel weighing and detection method according to claim 3, characterized in that: S1 also includes S13; S13. Preprocess the original weighing data sequence in the central industrial control processing module and extract the status dataset. The state dataset includes first-order derivative curve values. And the liquid density parameter p of the i-th channel i ; The preprocessing involves performing denoising and smoothing using a multi-point sliding window method based on the original weighing curve data W of each channel per unit time in the original weighing data sequence. Then, based on the denoised and smoothed original weighing curve data W of each channel per unit time, first-order difference calculations are performed to obtain the first-order derivative curve value. Where d represents the integral function, W i This represents the original weighing curve data of the i-th channel; simultaneously, based on the original weighing curve data W and the liquid volume, a density inversion calculation process is performed to extract the liquid density parameter p of the liquid in each channel.

5. The bus-type real-time multi-channel weighing detection method according to claim 4, characterized in that: S2 includes S21; S21. The first-order derivative curve value of each channel extracted from the state dataset. Extract the first derivative curve value within the set sampling window duration. The maximum and minimum values ​​in the sample are used to obtain the state drift index Py, which represents the degree of change in channel weighing, by dividing the difference between the maximum and minimum values ​​by the sampling window duration Δt.

6. The bus-type real-time multi-channel weighing and detection method according to claim 1, characterized in that: S2 further includes S22; S22. Based on historical multi-channel weighing data, the drift index Py of multiple channels in both stable and abnormal states is collected, and a drift threshold Pth is set. The drift threshold Pth is set by the upper limit of the stable state drift index Py. A preliminary comparison and evaluation is performed between the real-time acquired state drift index Py and the drift threshold Pth to determine the stability of the current channel weighing change. Based on the preliminary comparison and evaluation results, a dynamic sensitivity prediction mechanism is triggered. The specific evaluation content is as follows: When the state drift index Py ≥ drift threshold Pth, it indicates that the current channel weighing is unstable, and the dynamic sensitivity prediction mechanism is triggered at this time. When the state drift index Py < drift threshold Pth, it indicates that the channel weighing change is stable, and the current static and dynamic sensitivity correction process is adopted at this time.

7. The bus-type real-time multi-channel weighing detection method according to claim 6, characterized in that: S2 also includes S23; S23. After triggering the dynamic sensitivity prediction mechanism, the liquid density parameter p of the i-th channel is extracted from the industrial control processing module. i The sensitivity offset index ΔK is obtained by correlating the state drift index Py with the current channel sensitivity to measure the fluctuation difference between the current channel sensitivity and the neighboring channels.

8. The bus-type real-time multi-channel weighing detection method according to claim 6, characterized in that: S3 includes S31; S31. Based on the upper limit of the acceptable range of channel sensitivity deviation, set an offset threshold Kth, then compare and evaluate the real-time acquired sensitivity deviation index ΔK with the offset threshold Kth to determine the current sensitivity deviation of the weighing channel, and trigger the redundant channel compensation mechanism based on the results of the second comparison and evaluation; the specific evaluation content is as follows: When the sensitivity offset index ΔK > offset threshold Kth, it indicates an abnormal deviation, and the redundant channel compensation mechanism is triggered to correct the current channel weighing result. When the sensitivity offset index ΔK ≤ offset threshold Kth, it indicates that the deviation is normal, and the current channel weighing result should be maintained.

9. The bus-type real-time multi-channel weighing detection method according to claim 8, characterized in that: S3 also includes S31; S31. After triggering the redundant channel compensation mechanism, the original weighing value W of the i-th channel is used... i (t) is linearly fused with the weighted average weighing value Wneigh of the adjacent channel to obtain the compensated and corrected weighing value W'.

10. A bus-type real-time multi-channel weighing and detection system, applied to the bus-type real-time multi-channel weighing and detection method according to any one of claims 1-9, characterized in that: It includes a weighing sampling module, a weighing sensitivity analysis module, and a compensation analysis module; The weighing sampling module sets several sampling points in each weighing channel of the bus-type multi-channel weighing instrument to acquire the original weighing data sequence of each channel in real time, and inputs the original weighing data sequence to the industrial control processing module for extraction and preprocessing to extract the state dataset. The weighing sensitivity analysis module calculates the state drift index Py based on the state dataset, and sets a drift threshold Pth to perform a preliminary comparison and evaluation with the state drift index Py. If the preliminary comparison and evaluation indicates abnormal fluctuation, the dynamic sensitivity prediction mechanism is triggered to calculate and output the sensitivity offset index △K. The compensation analysis module performs a secondary comparison evaluation by setting an offset threshold Kth and a sensitivity offset index ΔK. If the sensitivity of the secondary comparison evaluation deviates abnormally, the redundant channel compensation mechanism is triggered.

Images