Solid powder intelligent weighing system and method based on fuzzy adaptive PID

Abstract

The invention discloses a solid powder intelligent weighing system based on fuzzy adaptive PID. The system comprises a material transfer module, a weighing unit, a material information identification module and an intelligent control unit. The material transfer module realizes automatic transfer and positioning of the weighing container; the weighing unit comprises a weighing container, a high-precision sensor and a discharging module, and the discharging module is provided with a material bin and a feeding valve mechanism; the information identification module is used for automatically identifying and inputting material information; the intelligent control unit is internally provided with a fuzzy self-adaptive PID controller, receives weight and material information, identifies valve-flow model parameters on line based on a recursive least square method and performs self-tuning, dynamically adjusts PID parameters according to a fuzzy rule, and outputs an instruction to realize three-stage control of rapid blanking, fuzzy fine adding and inching compensation. Through a parameter self-tuning and segmentation strategy, the method is adaptive to various material characteristics, the efficiency is improved and the static error is eliminated under the condition of ensuring high precision, and safe and traceable intelligent weighing is realized by relying on full-process automation and information tracing.

Description

Technical Field

[0001] This invention relates to the field of automated chemical experimental equipment technology, and more specifically, to a smart weighing system and method for solid powder based on fuzzy adaptive PID. Background Technology

[0002] In chemical synthesis experiments and production, high-precision weighing of solid powder materials is a core prerequisite for ensuring smooth reaction, stable product quality, and operational safety. However, this process currently relies heavily on manual operation or basic automated equipment, which presents significant limitations. Traditional manual weighing methods are not only inefficient but also prone to errors introduced by human factors, resulting in poor repeatability. Furthermore, operators must directly handle chemical reagents, facing health and safety risks.

[0003] To further improve efficiency and consistency, automated feeding equipment using vibratory or screw feeders has emerged in the market. These devices often employ simple on / off control or fixed-parameter PID control strategies. However, powdered materials generally possess complex physical characteristics such as nonlinearity, adhesion, and susceptibility to static electricity, resulting in significant time-varying behavior and diverse operating conditions. Fixed control parameters struggle to adapt to these dynamic changes, leading to problems such as low accuracy, severe overshoot, and a conflict between feeding efficiency and accuracy during the weighing process. Furthermore, they fail to effectively suppress dust, requiring frequent manual parameter adjustments and interventions, thus limiting the level of automation and intelligence.

[0004] Although some research has attempted to introduce advanced control algorithms such as fuzzy PID into batching systems, their applications are mostly concentrated in the continuous, high-volume material conveying of industries such as rubber and building materials. For the micro-volume, high-precision, intermittent point-like weighing requirements common in chemical synthesis, existing solutions lack targeted design. They often focus on adjusting the conveying speed, failing to achieve direct, closed-loop control of the final weighing accuracy. Furthermore, they fail to fully consider and address the challenges posed by the unique physical properties of solid powders, and their functions in system integration, material information management, and remote monitoring are also relatively lacking. Summary of the Invention

[0005] To address the aforementioned technical problems in related technologies, this invention proposes a smart solid powder weighing system and method based on fuzzy adaptive PID, which can overcome the above-mentioned shortcomings of the prior art.

[0006] To achieve the above-mentioned technical objectives, the technical solution of the present invention is implemented as follows: A smart solid powder weighing system based on fuzzy adaptive PID; The intelligent solid powder weighing system based on fuzzy adaptive PID includes a material transfer module, a weighing unit, a material information identification module, and an intelligent control unit. The material transfer module is used to realize the automatic grabbing, moving and positioning of weighing containers between the storage location and the weighing station; The weighing unit includes a weighing container, a high-precision weighing sensor, and a feeding module. The feeding module includes a material bin and a feeding valve mechanism. The material information recognition module is configured on the unloading module and is used to automatically identify and record material information; The intelligent control unit is signal-connected to the material transfer module, weighing unit and material information identification module, and includes an embedded fuzzy adaptive PID controller. The fuzzy adaptive PID controller is configured to: receive the weight feedback signal from the weighing sensor and the material information from the material information identification module; perform online identification and self-tuning of the system model parameters of "valve opening degree - feed flow rate" based on the recursive least squares method (RLS); dynamically adjust the PID control parameters in real time based on fuzzy inference rules, using the weight error (e) and error change rate (ec) as inputs; and output control commands to the feeding valve mechanism to realize three-stage weighing control of rapid feeding, fuzzy adaptive fine feeding, and jogging compensation.

[0007] Furthermore, the high-precision weighing sensor is a weighing module or analytical balance based on the principle of electromagnetic force compensation, the high-precision weighing sensor has a range of 320g, and the accuracy of the high-precision weighing sensor is not less than 0.0001g.

[0008] Furthermore, the feeding valve mechanism adopts a high-precision digital servo motor, the torque of the feeding valve mechanism is not less than 15 kg·cm, and the resolution of the feeding valve mechanism is not less than 0.1°.

[0009] Furthermore, the material information identification module is an RFID reader / writer device for reading material categories and formula parameters.

[0010] Furthermore, the main controller of the intelligent control unit is an STM32 or DSP series embedded processor.

[0011] According to another aspect of the present invention, a smart weighing method for solid powder is provided; The intelligent weighing method for solid powders includes the following steps: Step 1: Initialize system parameters and set the target weight value; Step 2: Rapid feeding stage, control the feeding valve mechanism to feed material at a large opening and large flow rate until the real-time weight reaches 85%~95% of the target weight; Step 3: Fuzzy Adaptive Fine Addition Stage. The fuzzy adaptive PID controller is started to calculate the weight error and error change rate in real time. The PID parameters are dynamically adjusted based on fuzzy rules, and the PID reference parameters are self-tuned online using the RLS algorithm. The valve opening is adjusted until the real-time weight reaches 98%~99.8% of the target weight. Step 4: In the jog compensation stage, the feeding valve mechanism is controlled to perform intermittent millisecond-level micro-motions to make up for the remaining weight with a small flow rate and eliminate static error; Step 5: Weighing complete, generate a traceable report.

[0012] Furthermore, in the fuzzy adaptive fine-tuning stage, the universe of discourse for the weight error e(k) is [-25, 25]mg, and the universe of discourse for the error change rate ec(k) is [-6, 6]mg / sampling period. Both are divided into 7 fuzzy subsets: {NB, NM, NS, ZO, PS, PM, PB}, and a triangular membership function is used.

[0013] Furthermore, the RLS algorithm employs a second-order ARX model with a forgetting factor, where the forgetting factor λ is set to 0.98~0.99, and the initial covariance matrix P(0) is set to 10. 6 ×I.

[0014] Furthermore, in the inching compensation stage, the pulse opening duration of the feeding valve mechanism is 20~100 ms, and the accuracy threshold δ is set to 0.05mg.

[0015] Furthermore, the switching threshold between the rapid feeding stage and the fuzzy adaptive fine machining stage is 90% of the target weight, and the switching threshold between the fuzzy adaptive fine machining stage and the jog compensation stage is 99% of the target weight.

[0016] The beneficial effects of this invention are as follows: By introducing a fuzzy adaptive PID controller and combining it with the recursive least squares method for online identification of system model parameters, real-time self-tuning and dynamic optimization of control parameters are achieved. This enables the system to automatically adapt to changes in the characteristics of different materials and fluctuations in operating conditions, effectively overcoming the shortcomings of poor adaptability of traditional fixed parameter controllers. Furthermore, by combining a three-stage segmented control strategy of rapid feeding, fuzzy adaptive fine feeding, and jogging compensation, the control mode is smoothly transitioned, thereby significantly improving the overall operating efficiency while ensuring extremely high weighing accuracy and completely eliminating steady-state errors. Finally, through the integration of fully automated operation and material information identification functions, operators can stay away from dangerous environments, achieving intelligent management with high safety, high robustness, and full-process data traceability in the weighing process. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a control logic diagram of a solid powder intelligent weighing system based on fuzzy adaptive PID according to an embodiment of the present invention; Figure 2 This is a flowchart of a smart solid powder weighing method based on fuzzy adaptive PID according to an embodiment of the present invention; Figure 3 This is a weight-time curve of the weighing process of a solid powder intelligent weighing system based on fuzzy adaptive PID according to an embodiment of the present invention. Figure 4 This is the ΔKp fuzzy rule table of a solid powder intelligent weighing system based on fuzzy adaptive PID according to an embodiment of the present invention; Figure 5 This is the ΔKi fuzzy rule table of a solid powder intelligent weighing system based on fuzzy adaptive PID according to an embodiment of the present invention; Figure 6 This is the ΔKd fuzzy rule table of a solid powder intelligent weighing system based on fuzzy adaptive PID according to an embodiment of the present invention. Detailed Implementation

[0019] 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. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention are within the scope of protection of the present invention.

[0020] It should be understood that in the description of the embodiments of the present invention, the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicating orientations or positional relationships, are based on the orientations or positional relationships shown in the accompanying drawings. They are used only for the convenience of describing the embodiments of the present invention and for simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the embodiments of the present invention. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more of the stated features. In the description of the embodiments of the present invention, "several" means two or more, unless otherwise explicitly specified.

[0021] like Figure 1 and Figures 3 to 6 As shown in the figure, a solid powder intelligent weighing system based on fuzzy adaptive PID according to an embodiment of the present invention comprises a material transfer module, a weighing unit, a material information identification module and an intelligent control unit.

[0022] The material transfer module is used to realize the automatic grabbing, movement and precise positioning of weighing containers between the storage location and the weighing station.

[0023] The weighing unit includes a weighing container for receiving materials, a high-precision weighing sensor for detecting the weight of the weighing container, and a feeding module. The feeding module includes a material bin and a feeding valve mechanism for controlling the powder flow rate, wherein the feeding valve mechanism may employ an actuator such as a high-precision servo motor.

[0024] The material information identification module is configured on the unloading module and is used to automatically identify and record the type and formula parameter information of the material during the material transfer process. For example, RFID technology can be used to realize the automatic collection and synchronous traceability of information.

[0025] The intelligent control unit, as the core control component of the system, includes a main controller with an embedded fuzzy adaptive PID controller. This unit is connected to the material transfer module, weighing unit, and material information identification module, and can interact with other automation platforms to achieve system linkage.

[0026] The fuzzy adaptive PID controller in the intelligent control unit is configured to perform the following functions: receive weight feedback signals from a high-precision weighing sensor; receive material information from a material information identification module; perform online identification and slow time-varying self-tuning of system model parameters between valve opening and feed flow rate based on a recursive least squares algorithm; dynamically adjust PID control parameters in real time based on fuzzy inference rules, with weight error and error change rate as inputs; and output control commands to the feeding valve mechanism to achieve three-stage weighing control: rapid feeding, fine feeding, and jogging compensation.

[0027] Secondly, such as Figure 2 As shown in the embodiment of the present invention, a smart weighing method for solid powder includes the following steps: First, the system is initialized, the controller parameters are initialized, and the target weight value is set. Then, the rapid feeding stage begins, where the controller controls the feeding valve mechanism to feed a large volume of material at a large opening until the real-time weight reaches the first threshold, such as 90% of the target weight. Next, the fuzzy adaptive fine-tuning stage begins. In this stage, the fuzzy adaptive PID controller is activated to collect weight data in real time and calculate the weight error and error change rate. The PID control parameters are dynamically adjusted based on the fuzzy rule base. At the same time, the baseline value of the PID parameters is self-tuned online based on the recursive least squares algorithm. The final PID parameters are synthesized and the control quantity is output to adjust the opening of the feed valve until the real-time weight reaches the second threshold, such as 99% of the target weight. Then, the inching compensation stage begins, controlling the feeding valve mechanism to perform intermittent millisecond-level micro-motions to replenish the remaining weight with a small flow rate, thus completely eliminating static error. Finally, after weighing is completed, the system automatically generates a traceable report.

[0028] To facilitate understanding of the above technical solutions of the present invention, the following detailed description of the above technical solutions of the present invention will be provided through specific embodiments.

[0029] In practical applications, the present invention provides a fuzzy adaptive PID-based intelligent weighing system and method for solid powders. Regarding the main hardware selection, the main controller can be an STM32F407 series, used for running the RLS algorithm and fuzzy inference. The high-precision weighing sensor can be a weighing module based on the electromagnetic force compensation principle or an analytical balance; in this embodiment, a weighing module with a range of 320g and an accuracy of 0.0001g is used. The weighing unit can employ a high-precision digital servo motor; in this embodiment, a high-precision digital servo motor with a torque of not less than 15kg·cm and a resolution of not less than 0.1° is used.

[0030] The software implementation includes the specific implementation and parameter selection of the fuzzy rule base, the specific implementation and parameter selection of the RLS algorithm, the description of the three-stage control algorithm, and the stage transition strategy and logical switching conditions.

[0031] The fuzzy adaptive PID controller takes the weight error e(k) and the error change rate ec(k) as inputs, and outputs the adjustment amounts ΔKp, ΔKi, and ΔKd of the proportional, integral, and derivative coefficients. The final PID parameters are synthesized by the following formulas: Kp = Kp0 + ΔKp, Ki = Ki0 + ΔKi, Kd = Kd0 + ΔKd, where Kp0, Ki0, and Kd0 are the reference values ​​of the PID parameters, provided by online tuning using the RLS algorithm.

[0032] In the input / output variable definitions, the weight error e(k) is defined as e(k) = W_target - W_current, where W_target is the target weight and W_current is the current weighing sensor reading. Its universe of discourse is [-25, 25] mg, which is set based on the starting point of the fine-tuning stage (e.g., 90% of the target weight) and allows for sufficient margin. The fuzzy subsets are divided into 7 categories: {Negative Large (NB), Negative Medium (NM), Negative Small (NS), Zero (ZO), Positive Small (PS), Positive Medium (PM), Positive Large (PB)}. The membership function uses a computationally efficient triangular function. A narrower distribution (base width approximately ±2mg) is used in the "Zero (ZO)" region to improve control accuracy near zero, while a wider distribution is used in the "Positive Large (PB) / Negative Large (NB)" regions to enhance the system's robustness to large deviations.

[0033] The error rate of change ec(k) is defined as ec(k) = e(k) - e(k-1). Its universe of discourse is [-6, 6] mg / sampling period, which is determined by the maximum feeding rate of the system and experimentally measured. The fuzzy subsets and membership functions are the same as e(k), dividing the system into 7 fuzzy subsets, using a triangular membership function.

[0034] The universe of discourse for the output variable PID parameter increments ΔKp, ΔKi, and ΔKd are ±50% of the initial reference values ​​(Kp0, Ki0, Kd0), i.e., [-0.5Kp0, 0.5Kp0], etc. This range ensures that parameter adjustments are sensitive without causing system instability. The fuzzy subsets are also divided into 7 subsets {NB, NM, NS, ZO, PS, PM, PB}, using a triangular membership function.

[0035] The fuzzy rule base is designed based on phase plane analysis of the step response of a second-order system and numerous simulation experiments. The rule form is: if e is A and ec is B, then ΔK is C. The following example illustrates the basis of the rule design.

[0036] For the fuzzy rule of ΔKp, for example, when e is NB (large negative error) and ec is NB (very rapid negative change), ΔKp is PB (significantly increasing proportional action), corresponding to a situation where the system is in a state of severe response lag, requiring the strongest control action to reverse the trend of error expansion. When e is NM (negative mean error) and ec is PS (slow positive change), ΔKp is PS (slightly increasing proportional action), corresponding to an improved system state but insufficient correction speed, requiring a moderately faster response. When e is ZO (zero error) and ec is ZO (zero rate of change), ΔKp is ZO (maintaining constant proportional action), corresponding to an ideal equilibrium state. The design basis is that ΔKp mainly affects the system response speed and is adjusted according to the magnitude of |e| and the trend of ec.

[0037] For the fuzzy rule ΔKi, for example, when e is NB and ec is NB, ΔKi is NB (significantly reducing the integral effect) to prevent integral saturation; when e is NM and ec is ZO, ΔKi is NS (slightly reducing the integral effect) to cautiously eliminate steady-state error; when e is ZO and ec is ZO, ΔKi is ZO (keeping the integral effect unchanged). The design basis is that ΔKi is mainly used to eliminate steady-state error, weakening the integral effect when |e| is large and strengthening the integral effect when |e| is small.

[0038] For the fuzzy rules of ΔKd, for example, when e is NB and ec is NB, ΔKd is PS (slightly increasing the derivative action) to suppress the error amplification trend; when e is NM and ec is PS, ΔKd is NS (slightly decreasing the derivative action) to avoid over-suppressing positive changes; when e is ZO and ec is ZO, ΔKd is NS (slightly decreasing the derivative action) to suppress high-frequency noise. The design is based on the fact that ΔKd is used to predict trends and suppress overshoot, and is adjusted according to the magnitude of |ec| and the system state.

[0039] The defuzzification method employs the centroid approach. The calculation formula is: the precise output value equals the sum of the products of the membership degrees of all output fuzzy sets and their centroid values, divided by the sum of the membership degrees of all output fuzzy sets. This method provides smooth output, high control precision, and is easy to implement in embedded controllers.

[0040] The RLS algorithm is used to identify the dynamic model parameters of a "valve opening - feed flow rate" system online, enabling slow time-varying self-tuning of PID baseline parameters. The system model employs a second-order autoregressive exogenous (ARX) model with a forgetting factor: y(k) = -a1·y(k-1) - a2·y(k-2) + b1·u(k-1) + b2·u(k-2) + ξ(k), where y(k) is the weight change rate, u(k) is the valve opening, ξ(k) is white noise, and a1, a2, b1, and b2 are the parameters to be identified. The model was chosen because a second-order model can capture the dynamic characteristics of the industrial process and has a moderate computational cost.

[0041] The steps of the RLS recursive algorithm include: Define the data vector: φ(k) = [-y(k-1), -y(k-2), u(k-1), u(k-2)]ᵀ; Define the parameter vector: θ(k) = [a1(k), a2(k), b1(k), b2(k)]ᵀ; Calculate the prior error: ε(k) = y(k) - φᵀ(k)·θ(k-1); Calculate the gain matrix: K(k) = P(k-1)·φ(k) / [λ+φᵀ(k)·P(k-1)·φ(k)]; Update parameter estimate: θ(k) = θ(k-1) + K(k)·ε(k); Update the covariance matrix: P(k)=[IK(k)·φᵀ(k)]·P(k-1) / λ.

[0042] In the selection of key parameters, the forgetting factor λ was set to 0.98 to achieve a balance between parameter tracking ability and estimation stability; the initial covariance matrix P(0) was set to 10. 6 ×I (I is the identity matrix) represents the uncertainty of the initial parameters, ensuring fast convergence; the initial parameter vector θ(0) can be set as a zero vector or a small positive vector.

[0043] During the PID reference parameter self-tuning process, every N sampling cycles (e.g., N=100), the controller recalculates the PID reference parameters (Kp0, Ki0, Kd0) using the system parameters θ(k)=[a1, a2, b1, b2] identified by the current RLS algorithm, through the pole placement method or an optimization algorithm based on integral performance indexes (e.g., ITSE). For example, the discrete model can be converted into a continuous transfer function to obtain the equivalent critical gain Ku and oscillation period Tu, and then the Ziegler-Nichols tuning formula can be used to calculate: Kp0=0.6Ku, Ki0=2Kp0 / Tu, Kd0=Kp0·Tu / 8.

[0044] The three-stage control algorithm includes a rapid feeding stage, a fuzzy adaptive fine-machining stage, and a jogging compensation stage. The rapid feeding stage aims to maximize efficiency by using a fixed high-power output (PWM duty cycle 100%) to feed material at full speed until the weight reaches the threshold η (e.g., 90%). This stage does not use a PID algorithm because maximizing output is the optimal strategy under large deviations.

[0045] The fuzzy adaptive fine-tuning stage establishes a hierarchical collaborative adaptive mechanism of fuzzy inference and RLS identification. The fast-response loop (upper layer - fuzzy adjustment) dynamically fine-tunes the PID parameters (ΔKp, ΔKi, ΔKd) based on the real-time error e and the error change rate ec in each control cycle (e.g., 50ms) to cope with nonlinearity, hysteresis, and instantaneous disturbances. The slow-adaptive loop (lower layer - RLS tuning) identifies the system model parameters online on a longer time scale and periodically retunes the PID reference parameters, enabling the controller to match the dynamic characteristics of the material and cope with slow time-varying disturbances. This "fast fuzzy, slow adaptive" structure achieves intelligent weighing.

[0046] The purpose of the jogging compensation stage is to eliminate steady-state error, employing rule-based on-off control. The algorithm is described as follows: if |e(k)|>δ, then issue an on pulse of fixed duration T_pulse (e.g., 50ms); otherwise, remain off. e(k) is the difference between the target weight and the current weight, and δ is the accuracy threshold (e.g., 0.05mg). The jogging mode avoids high-frequency oscillations through an incremental method, reliably eliminating steady-state error.

[0047] The phase transition strategy is based on multimodal intelligent decision-making. The switching condition from Phase 1 (rapid feeding) to Phase 2 (fuzzy adaptive fine-tuning) is that the real-time weight W_current >= W_target * η, where η is the rapid phase termination coefficient, which can be configured between 0.85 and 0.95 (preferably 0.90 in this embodiment). During the switching, mechanical inertia and material feeding in mid-air are taken into account to ensure overshoot prevention and smooth transition. The fuzzy adaptive PID controller seamlessly takes over from the current state.

[0048] The switching condition from stage two to stage three (jog compensation) is W_current>=W_target*μ, where μ is the termination coefficient of the fine-machining stage, which can be configured between 0.98 and 0.998 (preferably 0.98 in this embodiment). During the jog compensation stage, the feed valve is converted into a switching controller, outputting millisecond-level pulses (e.g., 20-100ms) until the error enters the dead zone (e.g., |e(k)|<0.05mg).

[0049] The switching condition for the completion of Phase 3 is |e(k)|<δ, where δ is the final accuracy index of the system (0.05mg in this embodiment).

[0050] During system initialization, the intelligent control system starts and self-checks the status of each component (such as high-precision servo motors, sensors, communication modules, and RFID devices). The material transfer module is in the initial position (grabbing standby state).

[0051] During the clamping, positioning, and information entry process, the material transfer module moves the weighing container to the weighing module (positioning error ≤ ±0.5mm), reading the RFID and entering the information during the process; the unloading module moves to directly above the weighing container (e.g., 5mm above the top plane of the weighing container), reading the RFID and entering the information during the process.

[0052] In the dynamic weighing example, the target weight is set to 200.0 mg. During the rapid feeding phase, stirring and a servo motor are activated for rapid feeding. When the real-time weight reaches approximately 180.0 mg (90%), the process transitions to the fuzzy adaptive fine-tuning phase. In this phase, the fuzzy adaptive PID controller is activated, and the servo motor adaptively fine-tunes. When the real-time weight reaches approximately 196.0 mg (99%), the process transitions to the jog compensation phase. During the jog compensation phase, the controller performs millisecond-level micro-motions, supplementing the weight with a small flow rate, ultimately reaching 200.0 mg.

[0053] Upon completion of weighing and reset, the material transfer module returns the unloading module after weighing and moves the weighing container to the next stage; the intelligent control system generates an experimental report (recording parameters such as weight settings, real-time weight, start time, and end time) and sends it to the remote terminal; the system resets to its initial state and waits for the next instruction.

[0054] In summary, by utilizing the technical solution of this invention, and by introducing a fuzzy adaptive PID controller combined with recursive least squares online identification of system model parameters, real-time self-tuning and dynamic optimization of control parameters are achieved. This enables the system to automatically adapt to changes in the characteristics of different materials and fluctuations in operating conditions, effectively overcoming the shortcomings of poor adaptability of traditional fixed-parameter controllers. Furthermore, by combining a three-stage segmented control strategy of rapid feeding, fuzzy adaptive precision feeding, and jogging compensation, the control mode is smoothly transitioned, thereby significantly improving overall operating efficiency while ensuring extremely high weighing accuracy and completely eliminating steady-state errors. Finally, through the integration of fully automated operation and material information identification functions, operators can stay away from hazardous environments, achieving intelligent management with high safety, high robustness, and full-process data traceability in the weighing process.

[0055] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A solid powder intelligent weighing system based on fuzzy adaptive PID, characterized in that, The material transfer module, the weighing unit, the material information identification module and the intelligent control unit are included. The material transfer module is used to realize automatic grabbing, moving and positioning of the weighing container between the storage position and the weighing position. The weighing unit includes a weighing container, a high-precision weighing sensor and a discharging module, and the discharging module includes a material bin and a feeding valve mechanism. The material information identification module is arranged on the discharging module and is used to automatically identify and input material information. The intelligent control unit is in signal connection with the material transfer module, the weighing unit and the material information identification module and includes an embedded fuzzy adaptive PID controller. The fuzzy adaptive PID controller is configured to receive weight feedback signals of the weighing sensor and material information of the material information identification module, to perform online identification and self-tuning on system model parameters of "valve opening-discharging flow" based on recursive least squares (RLS), to take weight error (e) and error change rate (ec) as inputs based on fuzzy inference rules, to dynamically adjust PID control parameters in real time, and to output control instructions to the feeding valve mechanism to realize three-stage weighing control of fast discharging, fuzzy adaptive precision adding and point-by-point compensation.

2. The solid powder intelligent weighing system based on fuzzy adaptive PID according to claim 1, characterized in that, The high-precision weighing sensor is a weighing module or an analytical balance based on electromagnetic force compensation principle, the range of the high-precision weighing sensor is 320 g, and the accuracy of the high-precision weighing sensor is not less than 0.0001 g.

3. The solid powder intelligent weighing system based on fuzzy adaptive PID according to claim 1, characterized in that, The feeding valve mechanism adopts a high-precision digital steering engine, the torque of the feeding valve mechanism is not less than 15 kg·cm, and the resolution of the feeding valve mechanism is not less than 0.1°.

4. The solid powder intelligent weighing system based on fuzzy adaptive PID according to claim 1, characterized in that, The material information identification module is an RFID read-write device for reading material categories and formula parameters.

5. The solid powder intelligent weighing system based on fuzzy adaptive PID according to claim 1, characterized in that, The main controller of the intelligent control unit is an STM32 or a DSP series embedded processor.

6. A solid powder smart weighing method using the system according to any one of claims 1 to 5, characterized in that, The method includes the following steps: Step one: initializing system parameters and setting a target weight value; Step two: fast discharging stage, controlling the feeding valve mechanism to discharge at a large flow rate with a large opening until the real-time weight reaches 85% to 95% of the target weight; Step three: fuzzy adaptive precision adding stage, starting the fuzzy adaptive PID controller, calculating weight error and error change rate in real time, dynamically adjusting PID parameters based on fuzzy rules, and adjusting the valve opening by using the RLS algorithm to online self-tune PID reference parameters until the real-time weight reaches 98% to 99.8% of the target weight; Step four: point-by-point compensation stage, controlling the feeding valve mechanism to perform intermittent millisecond-level micro-movement to supplement the remaining weight with a small flow rate and eliminate static error; Step five: weighing completion, generating a traceable report.

7. The method of claim 6, wherein, In the fuzzy adaptive precision adding stage, the universe of discourse of the weight error e(k) is [-25, 25] mg, the universe of discourse of the error change rate ec(k) is [-6, 6] mg / sampling period, both are divided into 7 fuzzy subsets: {NB, NM, NS, ZO, PS, PM, PB}, and a triangular membership function is used.

8. The method of claim 6, wherein the solid powder is a pharmaceutical powder. The RLS algorithm adopts a second-order ARX model with a forgetting factor, the forgetting factor λ is set to 0.98-0.99, and the initial covariance matrix P(0) is set to 10 6 ×I.

9. The method of claim 6, wherein, In the point-by-point compensation stage, the pulse opening duration of the feeding valve mechanism is 20 to 100 ms, and the accuracy threshold δ is set to 0.05 mg.

10. The method of claim 6, wherein, The switching threshold between the rapid discharging stage and the fuzzy adaptive fine adding stage is 90% of the target weight, and the switching threshold between the fuzzy adaptive fine adding stage and the point compensation stage is 99% of the target weight.

Images