An ink paste infusion system and a fault feedback design method thereof

Abstract

This invention discloses an ink paste filling system and its fault feedback design method. The ink paste filling system includes a metering pump with a motor connected to a transmission box. The transmission box is connected to a push rod located inside a cylinder. Driven by the motor, the transmission box drives the push rod to squeeze the ink paste inside the cylinder and fill it out through an output pipe. Multiple photocells, all connected to the cylinder, are used to detect the state of the ink paste inside the cylinder. An ink nozzle, connected to an input pipe, is used to inject ink paste into the cylinder through the input pipe. An ink paste cart includes a canister body with a bottom tank below it. The bottom tank is connected to a pressure pump to pressurize and deliver the ink paste from the bottom tank to the filling pipe and filling gun. The filling gun is connected to the ink nozzle to inject ink paste from the filling pipe into the ink nozzle. This application creates a local pressure difference inside the ink paste cart, squeezing out the ink paste inside the cavity. Simultaneously, it simulates potential faults that may occur during the operation of the metering pump and photocells, resulting in stronger system stability.

Description

Technical Field

[0001] This invention relates to the field of ink filling technology for cigarette machines, and particularly to an ink filling system and its fault feedback design method. Background Technology

[0002] During cigarette production, a special ink paste (thick type) is applied to the surface of the cigarette using a steel stamp. This stamp includes both brand and machine number stamps. Applying this ink paste not only increases brand recognition but also allows for quality traceability. The special ink paste needs to be supplied by an ink paste cart. The cart's filling gun injects the ink paste through the cigarette machine's filling nozzle into the ink paste metering pump. Once the metering pump is full, a photoelectric tube will display a message indicating it's full. Conversely, if the metering pump is low on ink paste, the photoelectric tube will display an empty pump message. Furthermore, during the ink paste injection process, if ink clumping or leakage is detected, affecting the ink paste flow rate and capacity, the photoelectric tube will detect these anomalies in real time and report fault information. The ink paste cart, as the ink paste supply terminal, continuously provides ink paste until the supply ends. Therefore, the nozzle, actuator, sensor, and ink paste cart are collectively referred to as the ink paste supply system (hereinafter referred to as the supply system).

[0003] There is a wealth of publicly available information on domestic grouting systems, such as the new single-pipeline grouting arm system developed by China Railway Chang'an Heavy Industry Co., Ltd. This system improves upon existing multi-pipeline design concepts by employing a grouting arm structure, thereby improving grouting quality, increasing construction efficiency, and reducing labor intensity. Anhui University of Technology used the MATLAB genetic algorithm toolbox to optimize the geometric dimensions of single piles, obtaining the optimal pile foundation design dimensions that meet both bearing capacity requirements and economic rationality, and verified this through examples. Shandong Jianzhu University conducted vertical and horizontal static load tests on pile foundations to study the effects of post-grouting technology on improving the soil layers around and at the pile tips, as well as enhancing the vertical and horizontal bearing capacity of the pile foundation. They proposed researching the influence of the pile-soil co-working effect and the horizontal resistance of the soil lateral to the pile cap on the horizontal bearing capacity of a single pile. Central South University studied the bearing capacity of screw piles using a hyperbolic model, empirical formula derivation, and grey system theory model. They derived a formula for determining the ultimate bearing capacity of screw piles. Based on measured data from static load tests of screw piles, they mathematically analyzed the Qs curve of the vertical static load test and compared it with the hyperbolic model and the derived ultimate bearing capacity to verify its feasibility and accuracy. Dalian University of Technology proposed an automatic injection system for porous structural insulation materials based on an industrial robot. Addressing the problems of poor injection height stability and low injection efficiency in manual injection, they designed an automatic injection device, studied a force-position coordinated injection control method, analyzed injection process parameters, and built a robotic automatic injection system. This achieved controllable injection force and trajectory, ensuring injection quality.

[0004] The composition and uses of the ink paste dispensing systems described above are different. This invention is also an ink paste dispensing system, but the contents and uses of this system are different from those disclosed above. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide an ink paste filling system and its fault feedback design method. The system uses a roller spiral propulsion method to create a local pressure difference in the filling chamber inside the ink paste carriage, thereby squeezing out the ink paste in the chamber. During this process, the faults that may occur during the operation of the metering pump and phototube are simulated and designed, and a fault feedback design method is designed for the above two.

[0006] The technical problem to be solved by the present invention is achieved through the following technical solution:

[0007] An ink paste dispensing system, comprising:

[0008] A metering pump includes a motor connected to a transmission box to drive the transmission box to move. The transmission box is connected to a push rod disposed in the cylinder. The cylinder is also provided with an input pipe and an output pipe. Driven by the motor, the transmission box further drives the push rod to squeeze the ink paste in the cylinder and pump it out through the output pipe.

[0009] Multiple photocells are connected to the cylinder body to detect the state of ink paste inside the cylinder body;

[0010] An ink nozzle, which is connected to the end of the input tube away from the cylinder, is used to inject ink paste into the cylinder through the input tube;

[0011] The ink cartridge truck includes a canister body, with a bottom tank connected to the bottom of the canister body. The bottom tank is connected to a pressure pump, which pressurizes the ink cartridge delivered from the bottom tank and delivers it to a filling tube and a filling gun located at the front end of the filling tube. The filling gun is connected to an ink nozzle and is used to inject the ink cartridge from the filling tube into the ink nozzle.

[0012] Preferably, the metering pump is located inside the cigarette machine SE supply assembly and is used for inputting and outputting ink paste; the phototube uses a high-precision optical sensor, and the push rod is made of stainless steel.

[0013] Preferably, the input pipe is detachably connected to the nozzle for inputting ink paste into the cylinder; the output pipe is detachably connected to the cigarette machine ink paste application device for outputting the ink paste from the cylinder.

[0014] Preferably, the nozzle is made of all-copper material, has a round opening, and is detachably connected to the infusion gun.

[0015] Preferably, a top cover is provided on the top of the tank body, and the two are connected by threads and are both made of thickened cylindrical aluminum. The top cover is provided with a handle and a pressure gauge, and the pressure gauge has a stroke of 0 to 5*10. 5 Pa is used to detect the pressure of compressed air inside the tank; the pressurization pump adopts a suction structure with a stroke of 0 to 0.25 MPa.

[0016] Preferably, the infusion system further includes a central electronic control system for controlling the operation of the infusion system, and the ink cartridge further includes a trolley with wheels at the bottom.

[0017] A fault feedback design method for an ink paste dispensing system is disclosed. This method simulates and constructs potential faults that may occur during the operation of the metering pump and phototube. It includes single-sided fault feedback design for the metering pump, single-sided fault feedback design for the phototube, and joint fault design for both the metering pump and phototube. Specifically, it includes the following steps:

[0018] S1. First, describe the working status of the infusion system. Then, by observing whether the input signal of the infusion system is present or absent over a period of time, determine whether the infusion system is faulty. When the infusion system is faulty, the metering pump, which is the actuator, and the phototube, which is the sensor, may both fail. It is necessary to consider how to effectively feed back the fault status. Therefore, it is necessary to design a fault feedback for the metering pump and the phototube.

[0019] S2. Metering pump one-sided fault feedback design: First, establish the static state value of the metering pump to obtain the static feedback controller; then select a suitable static feedback controller based on the static state value, i.e., a static positive / negative feedback controller; finally, design the metering pump one-sided fault feedback design model.

[0020] S3. Phototube One-sided Fault Feedback Design: First, establish the dynamic state value of the phototube to obtain the dynamic feedback controller; then, based on the fault response characteristics of the phototube, design two parameters to fit and correct the dynamic state value so that the output feedback closed loop of the phototube will not close quickly but will close with a delay. It is precisely this delay time that allows the infusion system to receive the fault information; finally, design the phototube one-sided fault feedback design model.

[0021] S4. Fault Design for Metering Pump and Phototube: First, the input signal (i.e., input vector) is used as the correlation factor to correlate the static and dynamic feedback controllers (regardless of the order of faults); then, after substituting the assignment results of the static and dynamic state values, the metering pump response coefficient and phototube gain coefficient no longer need to be involved in the calculation; finally, the output signal (i.e., output vector) is given in a unified manner.

[0022] Preferably, the method for determining whether the infusion system is faulty in S1 is as follows:

[0023] Establish the state feedback relation as follows: E(t)=f(x(t),y(t)), where t∈R, t represents unit time, R represents a real number, x(t)∈R, y(t)∈R, x(t) and y(t) represent the state feedback input vector (hereinafter referred to as input vector) and the state feedback output vector (hereinafter referred to as output vector) respectively, E(t)∈R, E(t) represents the state feedback quantity and is a matrix, and f represents the functional relationship;

[0024] The state feedback description for a continuous time period is represented as follows: The rank of E(t) satisfies rank(E) = R n-1 , where n represents a natural number and rank represents the rank of the matrix;

[0025] In the formula, when x(t) = 0, the irrigation system will definitely malfunction.

[0026] Preferably, step S2 is as follows:

[0027] Let the static state value be u(t), and its relationship be u(t) = (Ky(t)), where K represents the response coefficient of the metering pump and K∈R This leads to the static feedback controller model, namely: In the formula, z(t) represents the static feedback controller;

[0028] In the above formula, if u(t)≥0 then z(t)≥0, then z(t) is called a static positive feedback controller; conversely, if u(t)<0 then z(t)<0, then z(t) is called a static negative feedback controller.

[0029] When u(t) ≥ 0 or u(t) < 0, the corresponding z(t) is either a static positive feedback controller or a static negative feedback controller. The static state value u(t) is obtained by detecting the metering pump status signal through the program. Generally, this signal is a hexadecimal number. If the hexadecimal number is all positive, a static positive feedback controller is used (z(t) ≥ 0). If the hexadecimal number contains both positive and negative numbers, the time interval t from positive to negative or from negative to positive should be observed. The time interval t from negative to positive is the time to use a static negative feedback controller (z(t) < 0). In actual working conditions, the appropriate feedback controller can be selected for correct and effective fault feedback by detecting the metering pump status.

[0030] The design model for the one-sided fault feedback of the metering pump is as follows: In the formula, the static feedback control matrix Approximate correlation input vector matrix With static state matrix The result, and by the static feedback control matrix Feedback on the fault results.

[0031] Preferably, step S3 is:

[0032] When the phototube is functioning correctly, its gain varies from 0.1 to 0.3. However, when the phototube malfunctions, its gain value far exceeds the limit of 0.3. Based on the gain variation pattern during phototube malfunction, a dynamic state value u'(t) is established, i.e. In the formula, Q represents the phototube gain coefficient, and Q∈R, e represents a constant of 2.72, thus obtaining the dynamic feedback controller z'(t), that is: z'(t)=E -1 (t)f(u'(t)Q),x(t))=E -1 (t)f(u'(t)Q),E -1 (t)f(x(t));

[0033] For the dynamic state value u'(t), two parameters are used for fitting and correction: the dynamic input feedback coefficient C(t) and the dynamic output feedback coefficient D(t), with the following relationship: And D(t)=-ln(Q / C(t)), in this relationship, when C(t) decreases rapidly, D(t) initially increases, but the rate of increase will become slower and slower until the rate of increase cannot keep up with the rate of decrease of C(t). Therefore, the final result of u'(t) still shows a downward suppression phenomenon. So the final calculation result makes u'(t) still decrease, but the rate of decrease of u'(t) will not be fast.

[0034] A sufficient condition for the validity of the dynamic state value u'(t) is: for a phototube closed-loop system, if there are dynamic input feedback coefficients C(t) and dynamic output feedback coefficients D(t), then the dynamic state value u'(t) must have a positive definite matrix, i.e.:

[0035] In the formula, I represents the identity matrix. If this matrix has a feasible solution, then the system must have a dynamic feedback control quantity u'(t) for detecting phototube faults. When u'(t) ≥ 0, then z'(t) ≥ 0, and z'(t) is called a dynamic positive feedback controller. Conversely, when u'(t) < 0, then z'(t) < 0, and z'(t) is called a dynamic negative feedback controller. Therefore, the design model for single-sided fault feedback of the phototube is: In the formula, the dynamic feedback control matrix Approximate correlation input vector matrix With input feedback coefficient matrix Added output feedback coefficient matrix The result is determined by the dynamic feedback control matrix. Feedback on the fault results.

[0036] Preferably, step S4 is as follows:

[0037] First, the input vector is used as a correlation factor to correlate the static and dynamic feedback controllers. Then, the results of the static and dynamic state values ​​are substituted and transformed, eliminating the need for the metering pump response coefficient and phototube gain coefficient. Finally, the output signal is given in a unified manner.

[0038] There are static and dynamic feedback controllers corresponding to the metering pump and the phototube, namely:

[0039]

[0040] There are corresponding static and dynamic state values, namely:

[0041]

[0042] At this point, treating x(t) as a correlation factor transforms equation (1) into a static and dynamic correlation equation, namely:

[0043]

[0044] Substituting equation (2) into equation (3), we get:

[0045]

[0046] Through the evolved static and dynamic correlation equation (4), it can be observed that when x(t) is used as the correlation factor, both the static feedback controller z(t) and the dynamic feedback controller z'(t) have feasible solutions. Furthermore, the metering pump response coefficient and phototube gain coefficient do not need to be included in the calculation.

[0047] The above-described technical solution of the present invention has the following beneficial effects:

[0048] (1) A single-sided fault feedback of the metering pump was simulated and a static feedback controller was introduced to simulate the state feedback when the metering pump fails, thus avoiding abnormal phenomena such as invalid state feedback and untimely feedback.

[0049] (2) A single-sided fault feedback of the phototube was simulated and a dynamic feedback controller was introduced to simulate the state feedback when the phototube fails, thus avoiding abnormal phenomena such as invalid state feedback and untimely feedback.

[0050] (3) A simulation design was designed for the common failure of the metering pump and the phototube. When multiple components fail, regardless of the order in which the metering pump and the phototube fail, there is a corresponding closed-loop response and unified output information.

[0051] (4) The ink paste filling system of this application forms a local pressure difference inside the ink paste carriage to squeeze out the ink paste in the cavity. The system structure is ingenious and durable, and the operation is quick. Attached Figure Description

[0052] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments of the invention and, together with their description, serve to explain the principles of the invention.

[0053] Figure 1 This is a schematic diagram of the overall structure of the ink paste dispensing system of this application.

[0054] Figure 2 This is a partial schematic diagram of a metering pump.

[0055] Figure 3 This is a partial schematic diagram of the nozzle.

[0056] Figure 4 This is a partial schematic diagram of an ink paste printing press.

[0057] Figure 5 A flowchart for implementing the fault feedback design method.

[0058] Figure 6 The graph shows the trends of C(t) and D(t).

[0059] Figure 7 This is a trend graph of u'(t). Detailed Implementation

[0060] Various exemplary embodiments of the present invention will now be described in detail. It should be noted that, unless otherwise specifically stated, the relative arrangement, numerical expressions, and values ​​of the components and steps set forth in these embodiments do not limit the scope of the invention.

[0061] This application includes an ink paste dispensing system and its fault feedback design method. The following details the system with reference to specific embodiments:

[0062] Example 1: Ink Paste Infusion System

[0063] like Figure 1 , Figure 2 , Figure 3 and Figure 4 As shown, the ink paste dispensing system consists of a metering pump 1, a phototube 14, an ink nozzle 2, and an ink paste cart 3.

[0064] Metering pump 1 is installed inside the cigarette machine SE supply assembly for inputting and outputting ink paste. Metering pump 1 includes a motor 12, a transmission box 11, a cylinder 13, a photocell 14, a push rod 15, an input pipe 16, and an output pipe 17.

[0065] Furthermore, the motor 12 drives the transmission box 11 to move, and the transmission box 11 drives the push rod 15 to squeeze the ink paste in the cylinder 13 outward; when ink paste needs to be injected, the ink paste carriage injects it into the cylinder 13 through the ink nozzle and the input pipe 17.

[0066] Furthermore, the push rod 15 is located inside the cylinder body 13, is made of stainless steel, and is driven by the transmission box 11. Of course, this application is not limited to this, and those skilled in the art can choose a suitable push rod material as needed.

[0067] Furthermore, the input pipe 17 is connected to the ink nozzle to input ink paste into the cylinder 13; the output pipe 16 is connected to the cigarette machine ink paste application device to output the ink paste from the cylinder 13.

[0068] The phototube 14 uses a high-precision optical sensor, which is distributed on the cylinder 13 of the metering pump 1 to detect the ink paste status.

[0069] The nozzle 2 is made entirely of copper, with a round opening and a threaded end, and can be installed near the metering pump of a cigarette machine to connect to the metering pump 1. Of course, this application is not limited to this, and those skilled in the art can choose appropriate nozzle materials and shapes as needed.

[0070] The ink cartridge 3 consists of a top cover 11, a handle 33, a pressure gauge 32, a can body 34, a bottom can 35, a pressure pump 36, a filling pipe 37, a filling gun 38, and a trolley 39. As a separate component, when the cigarette machine needs ink, the ink cartridge 3 is pushed to the vicinity of the cigarette machine and ink is injected into the metering pump 1 through the nozzle. The ink in the cartridge is then replenished manually.

[0071] Furthermore, the top cover 31 is made of thickened, round aluminum, and the bottom of the cover has threads to connect to the can body 34. When it is necessary to open the top cover 31, simply rotate the top cover to open it.

[0072] Furthermore, the pressure gauge 32 has a stroke of 0 to 5*10. 5 Pa is used to detect the pressure of compressed air inside the tank 34.

[0073] Furthermore, the can body 34 is made of thickened cylindrical aluminum material to store ink paste.

[0074] Furthermore, the bottom tank 35 is used to transport the ink paste that naturally settles at the bottom of the tank 34 to the pressure pump 36.

[0075] Furthermore, the pressure pump 36 adopts a suction structure with a stroke of 0 to 0.25 MPa, which is used to pressurize and transport the ink paste from the bottom tank 35 to the filling pipe 37, and then inject it into the nozzle 2 through the filling gun 38.

[0076] Furthermore, the trolley 39 has four wheels at the bottom, two of which are swivel wheels, allowing it to move freely.

[0077] The ink dispensing system of this application is controlled by a central electronic control system (not shown).

[0078] The method of using the ink paste dispensing system of this application is as follows:

[0079] (1) When the cigarette machine requires ink paste:

[0080] First, phototube 14 detects that the ink paste in cylinder 13 of metering pump 1 is insufficient. At this time, phototube 14 sends a message that cylinder 13 is empty to the central control system of the cigarette machine. The central control system controls the input pipe 16 of metering pump 1 to open. Then, the trolley 39 is moved to the vicinity of the cigarette machine, and the filling gun 38 is inserted into the nozzle 2. After the pressurizing pump 36 injects compressed air, the ink paste is pressurized and flows through the filling pipe 37 to the filling gun 38. The ink paste is then injected into cylinder 13 of metering pump 1 through the nozzle. Once the ink paste in cylinder 13 is sufficient... After a certain amount of ink paste is stored, the phototube 14 detects that a certain amount of ink paste is stored in the cylinder 13 and then sends information to the central control system of the cigarette machine. The central control system controls the input pipe 16 of the metering pump 1 to close, and at the same time, the connection between the filling gun 38 and the nozzle 2 is manually removed. The central control system controls the motor 12 to drive the transmission box 11 to move. The transmission box 11 then drives the push rod 15 in the cylinder 13 to move. Finally, the ink paste in the cylinder 13 is output to the ink paste application device of the cigarette machine through the output pipe 17.

[0081] (2) When it is necessary to replenish the ink in the ink cartridge, first observe the change in the value of the pressure gauge 32. When the value decreases (the gas volume in the can increases), it indicates that the ink storage capacity is insufficient and needs to be replenished. At this time, after gripping the handle 33, open the top cover 31 upwards and at the same time turn off the external compressed air of the pressure pump 3 so that the filling gun 38 stops filling the ink. Then pour the prepared ink into the can 34. Finally, press the top cover 31 down and observe the change in the value of the pressure gauge 32 (the gas volume in the can decreases, causing the value to increase).

[0082] Example 2: Fault Feedback Design Method

[0083] The implementation steps of the fault feedback design method are as follows:

[0084] S1. Fault feedback design methods include single-sided fault feedback design for the metering pump, single-sided fault feedback design for the phototube, and joint fault design for both the metering pump and the phototube. In the fault feedback design process, the operating state of the infusion system is first described. Then, by observing the presence or absence of input signals over a period of time, it is determined whether the infusion system is faulty. When a fault occurs in the infusion system, both the metering pump (actuator) and the phototube (sensor) may malfunction. It is necessary to consider how to effectively feedback the fault status; therefore, fault feedback design is required for both the metering pump and the phototube.

[0085] S2. The design of a single-sided fault feedback system for the metering pump begins by establishing static state values ​​for the pump, thus obtaining a static feedback controller. Then, a suitable static feedback controller, i.e., a static positive / negative feedback controller, is selected based on the static state values. Finally, a single-sided fault feedback design model for the metering pump is designed.

[0086] S3. The design of a one-sided fault feedback system for the phototube involves first establishing dynamic state values ​​for the phototube to obtain a dynamic feedback controller. Then, based on the fault response characteristics of the phototube, two parameters are designed to fit and correct the dynamic state values, ensuring that the phototube's output feedback loop does not close rapidly but rather with a delayed closure. This delay is precisely what allows the infusion system to receive the fault information. Finally, a one-sided fault feedback design model for the phototube is designed.

[0087] S4. The aforementioned design for the joint failure of the metering pump and phototube first uses the input signal (i.e., the input vector) as a correlation factor to correlate the static and dynamic feedback controllers (regardless of the order of failure). Then, after substituting the assigned static and dynamic state values ​​and performing an evolution, the metering pump response coefficient and phototube gain coefficient are no longer needed. Finally, a unified output signal (i.e., the output vector) is provided. In this way, regardless of the order in which the metering pump and phototube fail (because the input vector is omitted), there is a corresponding closed-loop response (because the static and dynamic feedback controllers still exist and are correlated) and a unified output information (because feasible solutions exist for both).

[0088] The implementation process of the fault feedback design method is as follows: Figure 5As shown, the fault feedback design method indirectly reflects the state of the infusion system by providing state feedback from the metering pump and phototube. The core idea is to first construct the infusion system state based on the real-time feedback results of the actuators and sensors during the ink filling process, and then use the current system state to perform real-time feedback control, thereby ensuring system controllability and reliability. The purpose is to simulate faults in the metering pump and phototube in advance, which not only reduces the risk of abnormal state feedback but also optimizes the performance of the infusion system based on the fault results. Within the infusion system, the metering pump and phototube represent the actuators and sensors. In this application, "actuator" refers to the metering pump, and "sensor" refers to the phototube. It should be noted that in this infusion system, there is one metering pump and several phototubes, so the corresponding number of metering pumps and phototubes is one and several, respectively. The reliability of the state feedback from the metering pump and phototube is one of the key factors directly affecting system stability. To avoid abnormal phenomena such as invalid or untimely state feedback, a fault feedback design method for the metering pump and phototube is necessary.

[0089] Specifically:

[0090] The aforementioned fault feedback design method, in the context of the infusion system, involves simulating potential faults that may occur during the operation of the metering pump and phototube. First, the operating state of the infusion system is described as E(t) = f(x(t), y(t)), where t ∈ R, t represents a unit of time, R represents a real number, x(t) ∈ R, y(t) ∈ R, x(t) and y(t) represent the state feedback input vector (hereinafter referred to as the input vector) and state feedback output vector (hereinafter referred to as the output vector), respectively, E(t) ∈ R, E(t) represents the state feedback quantity and is a matrix, and f represents a functional relationship. Specifically, the state feedback description within continuous time can be expressed as... The rank of E(t) satisfies rank(E) = R n-1 , where n represents a natural number and rank represents the rank of the matrix.

[0091] The aforementioned fault feedback design method states that, under normal operating conditions, the input vector x(t) ≠ 0 of the infusion system. If and only if E(t) is a positive definite matrix or x(t) ≠ 0, it can be inferred that the infusion system is fault-free. Therefore, it can be deduced that if and only if E(t) is a non-positive definite matrix or x(t) = 0, the infusion system will definitely fail. Since the state feedback relation E(t) = f(x(t), y(t)) can be transformed into x(t) = E - 1Therefore, in actual operation, the presence or absence of input signals in the infusion system over a period of time can be used to determine whether the infusion system is faulty. Generally speaking, a normal infusion system will generate a large number of input signals, but once the infusion system malfunctions and stops, the input signals will no longer exist. Therefore, for the corresponding input vector x(t), we have For the description of the working state of the metering pump, if the rank of E(t) is not full definite, then E(t) must be a positive definite matrix. Therefore, the irrigation system can determine whether there is a fault simply by observing whether the input vector x(t) is equal to 0 or not equal to 0. When x(t) = 0, the irrigation system must have a fault. At this time, the metering pump, which is the actuator, and the phototube, the sensor, may both fail. It is necessary to consider how to effectively feed back the fault status. Therefore, it is necessary to design a fault feedback system for the metering pump and the phototube.

[0092] (1) Metering pump one-sided fault feedback design: The design steps are as follows: First, establish the static state values ​​of the metering pump to obtain the static feedback controller. Then, select a suitable static feedback controller based on the static state values, i.e., a static positive / negative feedback controller. Finally, design the metering pump one-sided fault feedback design model.

[0093] The aforementioned one-sided fault feedback design for the metering pump sets a static state value u(t), whose relationship is u(t)=(Ky(t)), where K represents the metering pump response coefficient and K∈R. This leads to the static feedback controller model, namely: In the formula, z(t) represents the static feedback controller.

[0094] As described above, the metering pump's one-sided fault feedback design dictates that the irrigation system will inevitably fail when x(t) = 0. Therefore, in the static feedback controller model, f(x(t)) = 0. Consequently, for the static feedback controller z(t), its feedback result depends on... The value of z(t) indicates that z(t) exhibits two trends. In equations (1) and (2), K is always greater than 1, therefore The value of z(t) ultimately depends on u(t). It should be noted that in (1), if u(t)≥0, then z(t)≥0. At this time, z(t) is called a static positive feedback controller. Conversely, in (2), if u(t)<0, then z(t)<0. At this time, z(t) is called a static negative feedback controller.

[0095] As described above, the single-sided fault feedback design for the metering pump, when u(t) ≥ 0 or u(t) < 0, corresponds to a static positive feedback controller or a static negative feedback controller. The static state value u(t) can be obtained by detecting the metering pump state signal through a program. This signal is generally a hexadecimal number. If all hexadecimal numbers are positive, a static positive feedback controller (z(t) ≥ 0) is used. If the hexadecimal number contains both positive and negative numbers, the time interval t from positive to negative or from negative to positive should be observed. The time interval t from negative to positive is the time for a static negative feedback controller (z(t) < 0). In actual operation, the appropriate feedback controller can be selected for correct and effective fault feedback by detecting the metering pump state.

[0096] As described above, the single-sided fault feedback design for the metering pump involves selecting a suitable feedback controller for fault feedback by detecting the metering pump's status. Therefore, the single-sided fault feedback design model for the metering pump is as follows: In the formula, the static feedback control matrix Approximate correlation input vector matrix With static state matrix The result, and by the static feedback control matrix Feedback on the fault results.

[0097] In the aforementioned metering pump single-sided fault feedback design, the phototube is an important component of the infusion system. When the phototube fails, the infusion system will also be unable to work stably. Therefore, the fault feedback design method should also consider the abnormal state feedback phenomenon caused by the single-sided failure of the phototube.

[0098] 2) Phototube One-Sided Fault Feedback Design: Unlike the static feedback controller of the metering pump, the phototube uses a dynamic feedback controller, and the phototube fault mainly manifests as a change in the gain range. The design steps are as follows: First, establish the dynamic state value of the phototube to obtain the dynamic feedback controller. Then, based on the fault response characteristics of the phototube, design two parameters to fit and correct the dynamic state value, ensuring that the phototube's output feedback loop does not close rapidly but rather with a delay. It is precisely this delay that allows the infusion system to receive the fault information. Finally, design a one-sided fault feedback model for the phototube.

[0099] The aforementioned phototube single-sided fault feedback design, in the infusion system, when the phototube is fault-free, its gain variation range is 0.1 to 0.3, while when the phototube malfunctions, its gain value will far exceed the limit of 0.3. Utilizing the gain variation law of the phototube during faults, a dynamic state value u'(t) is established, i.e. In the formula, Q represents the phototube gain coefficient, and Q∈R, and e represents a constant of 2.72. The dynamic feedback controller z'(t) is then obtained, i.e.: z'(t)=E- 1 (t)f(u'(t)Q),x(t))=E- 1 (t)f(u'(t)Q),E- 1 (t)f(x(t)).

[0100] The aforementioned phototube single-sided fault feedback design, for the feedback closed loop obtained after the phototube malfunctions, because the response speed of the phototube is instantaneous, when the phototube malfunctions, the infusion system has not yet received the fault information before the feedback closed loop of the phototube is closed. At this time, the infusion system is still in an unstable state. Therefore, a fault delay feedback function needs to be added to the dynamic feedback controller z'(t).

[0101] The aforementioned phototube single-sided fault feedback design ensures that when the phototube malfunctions, the instantaneous value of the phototube gain will exceed the limit, i.e., Q >> 0.3 and Q → +∞. At this point, the fault response relationship is: Observing this relationship, when the gain value Q increases rapidly, in order to smooth out the limit value, the dynamic state value u'(t) can only be reduced rapidly. In actual working conditions, this means that the feedback loop of the phototube is quickly closed. Therefore, the design steps for the phototube dynamic feedback controller z'(t) are to add some parameters to the fault response relationship to replace u'(t) and achieve the purpose of smoothing out the limit value.

[0102] The aforementioned phototube one-sided fault feedback design uses two parameters for fitting and correction of the dynamic state value u'(t): the dynamic input feedback coefficient C(t) and the dynamic output feedback coefficient D(t), with the following relationship: Furthermore, D(t) = -ln(Q / C(t)). Observing this relationship, when C(t) decreases rapidly, D(t) initially increases, but the rate of increase slows down until it can no longer keep up with the rate of decrease of C(t). Therefore, the final result of u'(t) still shows a downward trend. The trend graphs of C(t) and D(t) are shown below. Figure 6 As shown, the trend of u'(t) is as follows Figure 7 As shown, the final calculation result makes u'(t) still decrease, but the rate of decrease of u'(t) will not be fast. Therefore, in actual working conditions, when the phototube fails, the output feedback closed loop of the phototube will not close quickly but will close with a delay. It is precisely this delay time that allows the infusion system to receive the fault information.

[0103] The sufficient condition for the validity of the dynamic state value u'(t) in the aforementioned phototube one-sided fault feedback design is: for the phototube closed-loop system, if there are dynamic input feedback coefficients C(t) and dynamic output feedback coefficients D(t), then the dynamic state value u'(t) must have a positive definite matrix, that is:

[0104] In the formula, I represents the identity matrix. If this matrix has a feasible solution, then the system must have a dynamic feedback control quantity u'(t) for detecting phototube faults. When u'(t) ≥ 0, then z'(t) ≥ 0, and z'(t) is called a dynamic positive feedback controller; conversely, when u'(t) < 0, then z'(t) < 0, and z'(t) is called a dynamic negative feedback controller. Therefore, the design model for single-sided fault feedback of the phototube is: In the formula, the dynamic feedback control matrix Approximate correlation input vector matrix With input feedback coefficient matrix Added output feedback coefficient matrix The result is determined by the dynamic feedback control matrix. Feedback on the fault results.

[0105] (3) Design for simultaneous failure of metering pump and phototube: In the infusion system, a single failure of the metering pump or the phototube will directly affect the stability of the system. From the perspective of reliability design, if only single failure is considered, the performance of the infusion system will degrade or even lose stability, and the true stability of the infusion system cannot be achieved. In order to enable the infusion system to withstand the problem of simultaneous failure of two types of components, some public information described in the background of this application provides a reliability feedback design method to withstand the simultaneous failure of two or more components. However, the conclusions given are very conservative, and some require special operating conditions, so they have a certain degree of one-sidedness.

[0106] Regarding the design for simultaneous failure of the metering pump and phototube, this patent provides a reliability feedback design method for an infusion system where both the metering pump and phototube components fail simultaneously. The design steps are as follows: First, the input vector is used as a correlation factor to correlate the static and dynamic feedback controllers (regardless of the order of failure). Then, the assigned static and dynamic state values ​​are substituted and transformed, eliminating the need for the metering pump response coefficient and phototube gain coefficient. Finally, a unified output signal (i.e., output vector) is provided.

[0107] There are static and dynamic feedback controllers corresponding to the metering pump and the phototube, namely:

[0108]

[0109] There are corresponding static and dynamic state values, namely:

[0110]

[0111] At this point, treating x(t) as a correlation factor transforms equation (1) into a static and dynamic correlation equation, namely:

[0112]

[0113] Substituting equation (2) into equation (3), we get:

[0114]

[0115] Through the evolved static and dynamic correlation equation (4), it can be observed that when x(t) is used as the correlation factor, both the static feedback controller z(t) and the dynamic feedback controller z'(t) have feasible solutions. Furthermore, the metering pump response coefficient and phototube gain coefficient do not need to be involved in the calculation. In practical terms, this means that regardless of the order in which the metering pump and phototube fail (because the input vector is omitted), there will be a corresponding closed-loop response (because the static and dynamic feedback controllers still exist and are interconnected) and a unified output information (because feasible solutions exist for both).

[0116] The aforementioned design for the common fault of the metering pump and phototube presents the following conclusions: (1) Although the metering pump and phototube in the infusion system each implement feedback control for their own unilateral faults, the lack of consideration for the continuity and correlation of component faults will affect the system's reliability and cause the system to lose stability. (2) If only the metering pump and phototube faults are considered unilaterally, it is impossible to resist the system degradation caused by unilateral faults, and the system will also lose stability. Therefore, based on the above two conclusions, the design for the common fault of the metering pump and phototube is necessary (that is, if only the feedback control of the two unilateral faults is considered, it cannot be continuous and stable, so a common fault feedback design must be introduced).

[0117] From a broad system perspective, the ink delivery system is essentially a controlled system. This is because information such as ink flow rate, ink capacity in the metering pump, the empty / full load status of the metering pump, and the state of the delivery process detected by the phototube are all real-time and must be controlled. For a controlled system, the reliability of its operational status feedback is a prerequisite for stable operation. In the context of the ink delivery system, this manifests as the reliable correlation between ink flow rate feedback, ink capacity feedback, and ink status feedback. However, in the ink delivery system, the states of these three cannot be directly fed back; they are reflected indirectly through the status feedback from the metering pump and phototube, which connect these three components. Therefore, the stability of the ink delivery system depends on the reliability of the status feedback from the metering pump and phototube within the system.

[0118] This application has the following advantages:

[0119] (1) The main body used in this invention is a thick cigarette ink paste, whose fluidity, viscosity, color difference and other process indicators are significantly different from those of the main body described in the prior art, such as cement and refrigerant.

[0120] (2) The fault feedback design method adopted in this invention simulates the faults of the metering pump and phototube in advance without changing the original structure of the system. In contrast, the existing technology describes the redesign of the original structure or the redesign of the original structure operation mode.

[0121] (3) The reliable controller design method adopted in this invention takes into account the problem that the state cannot be measured when any component has multiple faults. After designing the fault feedback design method, the operating state can be measured and fault feedback can be provided. In contrast, the existing technology only simulates and provides feedback on a single-sided fault of a certain component.

[0122] Although the present invention has been disclosed above with reference to embodiments, it is not intended to limit the present invention. Any person skilled in the art can make various different choices and modifications without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention is defined by the claims and their equivalents.

Claims

1. A method of failure feedback design for an infusion system, the method comprising: The fault feedback design method is to simulate and design the possible faults in the working process of the metering pump and the photoelectric tube. The infusion system comprises a metering pump (1), the metering pump (1) comprises a motor (12), the motor (12) is connected with a transmission box (11) to drive the transmission box (11) to act, the transmission box (11) is connected with a push rod (15) arranged in a cylinder body (13), the cylinder body (13) is further provided with an input pipe (16) and an output pipe (17), a plurality of photoelectric tubes (14) are connected with the cylinder body (13), an oil nozzle (2) is connected with one end of the input pipe (16) away from the cylinder body (13), and an ink paste vehicle (3) is connected with the oil nozzle (2) through an infusion gun (38). The fault feedback design method comprises one-sided fault feedback design of the metering pump, one-sided fault feedback design of the photoelectric tube and common fault feedback design of the metering pump and the photoelectric tube, and specifically comprises the following steps: S1, first, the working state of the infusion system is described, then whether the infusion system is faulty is judged by observing whether the input signal of the infusion system has or has not in a period of time, when the infusion system is faulty, the metering pump as an actuator and the photoelectric tube as a sensor may be faulty, how to effectively feedback the fault state needs to be considered, therefore, the fault feedback design of the metering pump and the photoelectric tube needs to be carried out; S2, one-sided fault feedback design of the metering pump: first, the static state value of the metering pump is established to obtain a static feedback controller; then, a suitable static feedback controller, i.e. a static positive / negative feedback controller, is selected according to the static state value; finally, a one-sided fault feedback design model of the metering pump is designed; S3, one-sided fault feedback design of the photoelectric tube: first, the dynamic state value of the photoelectric tube is established to obtain a dynamic feedback controller; then, two parameters are designed to fit and correct the dynamic state value according to the fault response characteristics of the photoelectric tube, so that the output feedback closed loop of the photoelectric tube will not be closed rapidly but delayed, and it is this delay time that enables the infusion system to receive fault information; finally, a one-sided fault feedback design model of the photoelectric tube is designed; S4, common fault design of the metering pump and the photoelectric tube: first, the input signal is taken as a correlation factor to correlate the static and dynamic feedback controllers; then, the assignment results of the static and dynamic state values are substituted into the evolution, so that the metering pump response coefficient and the photoelectric tube gain coefficient are no longer needed to participate in the calculation; finally, the output signal is given.

2. The fault feedback design method of an infusion system according to claim 1, characterized in that, The method for judging whether the infusion system is faulty in S1 is: The state feedback relation is established, that is, , wherein , t represents a unit time, R represents a real number, , x(t) and y(t) respectively represent a state feedback input vector and a state feedback output vector, , E(t) represents a state feedback amount and is a matrix, f represents a function relation; For the state feedback description in continuous time is expressed as , E(t) The rank of satisfies , n denotes a natural number, rank denotes the rank of a matrix; In the formula, when x(t)=0 the infusion system must be malfunctioning.

3. The fault feedback design method of an infusion system according to claim 2, wherein, Step S2 is: Setting a static state value u(t) , where , where K represents the metering pump response coefficient and , and then the static feedback controller model is obtained, i.e. , where z(t) represents the static feedback controller; In the above formula, if u(t)≥0 then z(t)≥0 at this time z(t) is called a static positive feedback controller, on the contrary, if u(t)<0 then z (t)<0 at this time z(t) is called a static negative feedback controller; In u(t)≥0 or u(t)<0 , its corresponding z(t) is a static positive feedback controller or a static negative feedback controller, static state value u(t) obtained by program detection metering pump state signal, generally the signal is hexadecimal number, if the hexadecimal number is all positive number, static positive feedback controller is adopted (z(t)≥0) , if the hexadecimal number has positive number and negative number, at this time, the time interval of positive number to negative number t or the time interval of negative number to positive number t , the latter from negative number to positive number t transformation time, static negative feedback controller is adopted (z(t)<0) , in actual working condition, suitable feedback controller can be selected through detection metering pump state to carry out correct and effective fault feedback, The single-sided failure feedback design model for a metering pump is: where the static feedback control matrix is approximated by the input vector matrix and the result of the static state matrix is fed back by the static feedback control matrix .

4. The fault feedback design method of an infusion system according to claim 2, wherein, Step S3 is: When the photocell is not faulty, its gain variation range is 0.1-0.3, and when the photocell is faulty, its gain value will be far more than the limit value of 0.

3. The gain variation rule of the faulty photocell is used to establish the dynamic state value u’ (t)， That is , wherein Q represents the gain coefficient of the photocell, and , e represents a constant of 2.72, and then the dynamic feedback controller z'(t) is obtained, that is ; For dynamic state value u’(t) , two parameters are used to fit the correction, respectively, dynamic input feedback coefficient C(t) and dynamic output feedback coefficient D(t) , there is a relationship , in which, when C(t) rapidly decreases, D(t) just start to increase, but the increase will be more and more slow, until the increase speed can not keep up with C (t) decrease speed, so eventually make u’(t) results still show the phenomenon of depression, so the final calculation results make u’(t) is still decreasing, but make u’(t) decrease speed will not be fast; dynamic state value u’(t) A sufficient condition for the photocell closed loop system to be effective is that if there exists a dynamic input feedback coefficient C(t) and a dynamic output feedback coefficient D(t) then the dynamic state value u’(t) must exist a positive definite matrix, that is: , where I denotes the identity matrix, if the matrix has a feasible solution, the system must have a dynamic feedback control quantity to detect the fault of the photocell u’(t) , and when u’(t) ≥0 , then z’(t)≥0 , then z’(t) is called a dynamic positive feedback controller; otherwise, when u’(t)<0 , then z’(t)<0 , at this time z’ (t) is called a dynamic negative feedback controller, so the unilateral fault feedback design model of the photocell is: where the dynamic feedback control matrix approximates the input vector matrix and the input feedback coefficient matrix and the output feedback coefficient matrix is the result of the addition of the dynamic feedback control matrix and the feedback fault result.

5. The fault feedback design method of an infusion system according to claim 2, wherein, Step S4 is: Firstly, the input vector is taken as the correlation factor, the static and dynamic feedback controllers are correlated, and then the assignment results of the static and dynamic state values are substituted and evolved, so that the response coefficient of the metering pump and the gain coefficient of the photoelectric tube are no longer needed, and finally the output signal is given; There are static and dynamic feedback controllers corresponding to the metering pump and the photoelectric tube, that is, ， There are static and dynamic state values corresponding to them, that is, ， At this time, the x(t) As the correlation factor, the formula (1) is evolved into the static and dynamic correlation formula, i.e.: , Substitute equation (2) into equation (3), that is, ， By the evolved static and dynamic correlation equation (4) can be observed when x(t) The static feedback controller z(t) And the dynamic feedback controller z'(t) Both exist feasible solution And no need to meter pump response coefficient and photoelectric tube gain coefficient involved in the calculation.

Images