Cooperative control method of photovoltaic intelligent manufacturing equipment line

Abstract

The present application relates to the technical field of control or regulation system, disclose a kind of photovoltaic intelligent manufacturing equipment production line's collaborative control method, comprising: control unit real-time acquisition physical buffer area in the material inventory and change rate;Establish and map physical buffer area as virtual viscoelastic dynamics model with non-newtonian fluid characteristics;Based on the model calculation makes inventory regression equilibrium point virtual elastic recovery force and virtual viscous damping force that impedes inventory state change;Wherein, execute asymmetric anisotropic damping generation strategy, according to material flow trend dynamic split damping coefficient;Finally, the virtual adjustment correction amount after vector synthesis is added to the basic transmission speed to generate dynamic speed command, the present application constructs the virtual dynamics field and asymmetric damping mechanism of rheological characteristics, solve the nonlinear cascade oscillation problem in discrete logistics transmission, realize the differentiated adaptive inhibition of accumulation and dispersion risk.

Description

TECHNICAL FIELD

[0001] The present application relates to a kind of photovoltaic intelligent manufacturing equipment line's cooperative control method, belong to control or regulating system technical field. BACKGROUND

[0002] Current photovoltaic cell and component precision manufacturing production line is usually composed of multiple process equipment through physical buffer zone in series, and the existing stage system control generally follows the single machine independent operation to cooperate with the physical buffer coupling discrete event driven mode, each process unit is independently operated according to its own set beat, and the material transfer between devices depends on the high and low threshold signals of the material inventory of the physical buffer zone, which triggers the start-stop or multi-section speed switching of the upstream equipment. This control strategy uses the logical decoupling characteristic to maintain the basic logistics transmission of the production line under low speed or low coupling condition. The industry tries to introduce a cooperative mechanism based on data analysis, for example, the Chinese patent application with publication number CN118536953A discloses a production line production cooperation method, system, electronic device and medium. The scheme matches the production task attributes with the preset over-production rate, combines the upstream task breakage rate and the excellent rate interval to determine the over-production reason, realizes the cooperation and correction at the production task level, and the essence of the scheme is based on the discrete statistical data macro management logic, focuses on production task accounting, post-tracing and state confirmation, and the static or quasi-static cooperation based on statistical interval cannot penetrate to the bottom of the device dynamics control. In the face of high-speed transmission of ultra-thin sheets in photovoltaic production line, the method cannot respond to the sudden change of logistics flow in milliseconds, lacks real-time suppression mechanism for physical inertia and mechanical shock, and the micro transmission level of the production line is still in a rigid confrontation state.

[0003] In the face of high-throughput, ultra-thin substrate and high requirements for continuous rheological properties in modern manufacturing environment, the discrete threshold rigid control paradigm has limitations in dynamics. The existing technology regards the continuous flow characteristic material flow as a discrete particle set, and uses finite state machine switch logic to constrain the fluid characteristic flow dynamics. This control method ignores the fluid characteristics and wave conduction characteristics presented by the stage system under high-speed running state. The existing technology has the following disadvantages: discrete control leads to nonlinear cascade oscillation, because there is a lag in the mechanical inertia and control response of the upstream equipment, simple start-stop or step speed instruction makes the logistics speed switch sharply between full load and starvation state, the system lacks internal damping mechanism to absorb and dissipate disturbance energy, and extreme limit cycle oscillation leads to frequent changes in transmission speed; disturbance amplification effect, small beat fluctuation of downstream process is amplified to a large speed adjustment of upstream equipment through physical buffer zone threshold quantization processing, so that the system is difficult to maintain global dynamic balance under long-chain coupling.

[0004] Therefore, how to break through the traditional discrete threshold control rigid constraint, and construct a flexible logical connection mechanism with adaptive buffer and energy dissipation capability by converting the rigid physical connection between devices into a flexible logical connection mechanism, to realize the continuous modulation and dynamic suppression of logistics fluctuation in multi-stage serial discrete manufacturing system, has become a technical problem to be solved by the present application. SUMMARY

[0005] To solve the problems presented in the background art, the technical solution of the present application is as follows: A photovoltaic intelligent manufacturing equipment production line cooperative control method, the production line includes multiple process units connected through physical buffer zones, the method maps the discrete material flow in the physical buffer zone into a virtual viscoelastic dynamics model with non-Newtonian fluid characteristics, and the method includes the following steps:

[0006] The control unit of each process unit collects the current material inventory in the associated physical buffer zone and the change rate of the material inventory in real time;

[0007] The control unit calculates the virtual elastic restoring force that returns the material inventory to the preset equilibrium point and the virtual viscous damping force that hinders the change of the material inventory state based on the virtual viscoelastic dynamics model;

[0008] Wherein, when calculating the virtual viscous damping force, the control unit performs an asymmetric anisotropic damping generation strategy: real-time identification of the positive and negative polarity of the change rate of the material inventory to determine the immediate trend of the material flow, when determining that the trend is material accumulation, a first damping coefficient is called to calculate an over-damping force with strong inhibition characteristics, when determining that the trend is material dispersion, a second damping coefficient smaller than the first damping coefficient is called to calculate an under-damping force with weak inhibition characteristics;

[0009] The control unit vectorially combines the virtual elastic restoring force and the virtual viscous damping force obtained based on the asymmetric anisotropic damping generation strategy to generate a virtual adjustment correction, and superimposes the virtual adjustment correction on the basic transmission speed of the process unit to generate a dynamic speed instruction for driving the transmission mechanism;

[0010] Through the asymmetric anisotropic damping generation strategy, the process unit's response to upstream flow fluctuations exhibits a nonlinear rheological characteristic that coexists with rigid inhibition under the accumulation trend and flexible following under the dispersion trend.

[0011] Preferably, the calculation logic of the virtual adjustment correction in the virtual viscoelastic dynamics model follows a linear combination relationship containing a state deviation term and a state change rate term, and satisfies the following dynamic equation: wherein, represents the virtual adjustment correction, represents the virtual elastic stiffness coefficient represents the current material inventory represents the preset equilibrium point, represents the flow trend, the first damping coefficient or the second damping coefficient is dynamically switched, represents the change rate of the material inventory; the equation defines the dynamic convergence trajectory of the system when deviating from the steady state.

[0012] Preferably, the method further comprises establishing a virtual back-pressure communication mechanism for transmitting in reverse direction along the material flow: the control unit receives a virtual pressure value sent by a directly downstream process unit, and introduces the virtual pressure value as a feed-forward suppression variable into the generation logic of the dynamic speed command; the virtual pressure value is generated by the downstream process unit based on its own current material inventory and the virtual pressure value it receives from a more downstream process unit through weighted accumulation calculation; through the virtual back-pressure communication mechanism, the material flow blockage state of the downstream is diffused in reverse direction to the upstream in the form of a scalar field, driving the upstream process unit to perform predictive deceleration before the physical buffer zone is full, and using the lead transmission of information flow to exchange for the redundancy of physical buffer space.

[0013] Preferably, the step of calculating the virtual elastic restoring force comprises executing a nonlinear boundary hardening strategy: the control unit monitors the safety margin between the material inventory and the physical limit boundary of the physical buffer zone in real time; the control unit generates a nonlinear gain coefficient based on the safety margin, which is set to remain a constant base value when the safety margin is greater than a preset threshold, and is monotonically increasing in a nonlinear manner when the safety margin is less than the preset threshold and tends to zero; the control unit uses the nonlinear gain coefficient to weight and amplify the virtual elastic restoring force, so as to generate an exponentially enhanced repulsive potential field when the material inventory approaches the physical limit boundary, so that the transmission mechanism continuously corrects the speed in the opposite direction without external hard limit triggering.

[0014] Preferably, the method further comprises executing a reference speed adaptive correction step: the control unit sets a sliding time window and calculates a statistical average value of the virtual adjustment correction within the window; when the absolute value of the statistical average value exceeds a preset dead zone threshold, the control unit identifies that there is a systematic speed mismatch, and compensates and corrects the base transmission speed based on the statistical average value, and uses the corrected base transmission speed as an updated reference value for the generation of subsequent dynamic speed commands; through the correction step, the steady-state inventory deviation caused by equipment aging or environmental drift is eliminated, so that the virtual adjustment correction automatically converges to the vicinity of zero point under steady-state operation, and the dynamic range of the system bidirectional adjustment is restored.

[0015] Preferably, the method further comprises executing a resonance active suppression step: the control unit monitors the sign flip frequency of the virtual adjustment correction in the time domain to estimate the real-time oscillation frequency of the cooperative control system; when the real-time oscillation frequency falls within a preset intrinsic resonance frequency band, the control unit applies a time-varying offset to the virtual elastic stiffness coefficient in the virtual viscoelastic dynamic model, dynamically changes the natural frequency of the cooperative control system to destroy the resonance phase synchronization condition; the control unit temporarily increases the weight of the virtual viscous damping force when the resonance trend is detected, to dissipate the resonance energy in the system.

[0016] Preferably, the method further comprises a step of inertia-adaptive gain scheduling based on the material inventory: the control unit maps the material inventory in real time into an inertia factor representing the current load inertia of the system; the control unit dynamically adjusts the virtual elastic stiffness coefficient and the virtual damping coefficient in the virtual viscoelastic dynamic model in real time based on the inertia factor and according to a preset nonlinear mapping rule; the nonlinear mapping rule is set to correspondingly increase the virtual elastic stiffness coefficient and the virtual damping coefficient with the increase of the inertia factor, so as to maintain the natural frequency and the damping ratio of the coordinated control system constant under different load conditions and ensure the consistency of the dynamic response characteristics of the system.

[0017] Preferably, the acquisition of the change rate of the material inventory in the state acquisition step comprises: acquiring the difference of the material inventory between two adjacent sampling periods; performing low-pass filtering processing on the difference of the material inventory to filter out high-frequency noise interference and generate a smoothed change rate signal as the basis for calculating the virtual viscous damping force; the cutoff frequency of the low-pass filtering processing is set to be lower than the lowest effective response frequency of the coordinated control system, so as to prevent the asymmetric anisotropic damping generation strategy from being triggered by false sensor signal jitter.

[0018] Preferably, the step of generating a dynamic speed instruction further comprises nonlinear saturation constraint processing: the control unit presets a speed constraint range based on the minimum residence time or the maximum physical beat limit of the process step performed by the process unit; when the calculated dynamic speed instruction exceeds the speed constraint range, the control unit clamps the dynamic speed instruction at the boundary value of the speed constraint range, to preferentially guarantee that the single-machine process parameters do not exceed the limit.

[0019] Preferably, the step of adaptively correcting the reference speed further comprises state freezing logic: the control unit monitors the running state of the production line in real time; when it is detected that the production line is in a non-steady-state condition such as start-up, emergency stop or fault alarm, the control unit automatically suspends the step of adaptively correcting the reference speed, keeps the current basic transmission speed unchanged, and prevents transient disturbances from being incorrectly integrated into the reference value.

[0020] Compared with the prior art, the present application has the following advantages:

[0021] 1. In photovoltaic intelligent manufacturing equipment, a virtual viscoelastic impedance model is constructed between adjacent process units, the material inventory and change rate of the downstream physical buffer area are mapped in real time as virtual elastic restoring force and virtual damping inhibition force in the control loop, the mechanism changes the traditional discrete manufacturing based on threshold switch rigid control mode, based on the continuous fluid rheology characteristics of discrete logistics, the virtual damping component is calculated by the material inventory change rate, the energy dissipation mechanism for logistics fluctuation is introduced in the control algorithm layer, when the upstream transmission beat occurs transient disturbance, the damping mechanism produces nonlinear resistance according to the disturbance intensity, actively absorbs and attenuates the shock energy, so that the system state does not need to be mechanically stopped or intervened, relies on the convergence characteristics of the control law to smoothly return to the steady state equilibrium point, and eliminates the nonlinear cascade shock phenomenon commonly found in long-chain production lines.

[0022] 2. A virtual back pressure communication mechanism is established in the reverse direction along the material flow direction, so that the load pressure of the downstream process unit penetrates upstream in the form of a scalar field, the information flow and the material flow are reversely coupled, so that the upstream process unit is no longer blindly running, the back pressure signal is received to sense the remote downstream blocking trend in advance, which is introduced as a feedforward inhibition variable into the current speed generation logic, the physical space redundancy is obtained by using the information advance, and the upstream equipment is driven to perform predictive speed adjustment before the physical congestion actually arrives, so that the local severe stacking impact that may occur is converted into a small speed coordination of all devices on the whole line, and distributed absorption and spatial attenuation of flow fluctuation are realized.

[0023] 3. When calculating the virtual damping inhibition component, an asymmetric anisotropic strategy is adopted, the system damping characteristics are dynamically reconstructed according to the material inventory change trend direction, when the buffer area is in the filling trend, the system automatically adapts to the high damping coefficient to build the over-damping characteristics, provides strong virtual brake force to preferentially guarantee physical safety, and prevents material collision; when the buffer area is in the emptying trend, the system automatically switches to the low damping coefficient to build the under-damping characteristics, allows the virtual elastic force to dominate to realize fast replenishment response, automatically switches the control parameter logic according to the flow direction polarity, solves the inherent contradiction between the steady-state accuracy and the dynamic response speed of the traditional linear control in a single control loop, ensures that the production line has high anti-blocking and starvation elimination ability, and maximizes the system dynamic bandwidth. BRIEF DESCRIPTION OF DRAWINGS

[0024] Fig. 1 The present application is a collaborative control flowchart integrating asymmetric damping and virtual back pressure regulation;

[0025] Fig. 2 The present application is a virtual force and inventory change rate waveform diagram showing asymmetric anisotropic characteristics;

[0026] Fig. 3 The present application is a control system architecture diagram of a multi-dimensional inhibition and self-adaptive correction module. DETAILED DESCRIPTION

[0027] In order to better understand the technical scheme of the present application, the technical scheme in the embodiments of the present application is described clearly and completely below in combination with the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work should fall within the scope of the present application.

[0028] A photovoltaic intelligent manufacturing equipment production line collaborative control method runs on a distributed control architecture based on an industrial field bus, wherein the field bus can adopt EtherCAT or Profinet communication protocol, contains multiple process units connected in series according to a process flow, adjacent process units are coupled through a physical buffer area, each process unit is equipped with an independent control unit and a sensor assembly for monitoring the feeding and discharging state, the method establishes the mapping relationship between the discrete material flow in the physical buffer area and the virtual viscoelasticity dynamic model through the collaboration of each functional module, and realizes the continuous collaborative adjustment of the production line flow; the control unit of each process unit collects the current material inventory in the associated physical buffer area and the change rate of the material inventory at a preset control period , wherein the control period can be set to 2ms to 10ms, for the acquisition of the change rate of the material inventory , the control unit collects the difference of the material inventory of the adjacent two sampling periods, and applies low-pass filtering processing to the difference sequence, the low-pass filtering processing adopts a first-order lag filtering algorithm or a moving average filtering algorithm, the cutoff frequency is set to be lower than the lowest effective response frequency of the collaborative control system, for example, 5Hz to 10Hz, to filter out the high-frequency noise components caused by sensor jitter, and generate a smoothed change rate signal; the control unit establishes the mapping of the physical buffer area and the virtual viscoelasticity dynamic model with non-Newtonian fluid characteristics, and calculates a virtual adjustment correction amount for correcting the transmission speed based on the model in real time, the calculation logic follows a linear combination relationship containing a state deviation term and a state change rate term, and satisfies the following dynamic equation: , wherein, represents a preset equilibrium point, and the value is set to 50% of the capacity of the physical buffer area; represents a virtual elastic stiffness coefficient, used to generate a virtual elastic restoring force to return the material inventory to the equilibrium point; represents a damping coefficient that dynamically switches with the flow trend, used to generate a virtual viscous damping force to hinder the change of the material inventory state.

[0029] In calculating the virtual viscous damping force, the control unit performs an asymmetric anisotropic damping generation strategy, and the control unit identifies the change rate of the material inventory in real time The positive or negative polarity is used to determine the immediate trend of material flow. When detected When the value is greater than zero, indicating a trend of material accumulation, the control unit calls the first damping coefficient. Calculate the virtual viscous damping force; the first damping coefficient The numerical setting is used to construct the overdamped control characteristics to generate a virtual resistance force that inhibits material stacking when detected. When the value is less than zero, indicating a trend of material evacuation, the control unit activates the second damping coefficient. Calculate the virtual viscous damping force; the second damping coefficient The value is less than the first damping coefficient. This is used to construct underdamped control characteristics, allowing virtual elastic restoring force to dominate the regulation process; the control unit executes a virtual back pressure communication mechanism that transmits pressure in the reverse direction of the material flow along the production line, and receives virtual pressure values ​​sent by the direct downstream process unit via the industrial bus. This value is then used as a feedforward suppression variable in the dynamic speed command generation logic, and the local virtual pressure value is... The downstream process unit is determined based on its current material inventory. The received downstream virtual pressure values ​​are generated through the following weighted accumulation logic: ,in, This is the local load weighting coefficient. This is the pressure transmission attenuation coefficient, with a value ranging from 0 to 1; To determine the physical buffer's maximum capacity, this mechanism drives upstream process units to detect remote congestion and implement deceleration adjustments before the physical buffer reaches its limit. The control unit executes a nonlinear boundary hardening strategy to calculate virtual elastic restoring force, and monitors material inventory in real time. Safety margin between physical buffer limits and physical buffer boundaries Based on this safety margin The control unit uses hyperbolic or exponential functions to generate nonlinear gain coefficients. The nonlinear gain coefficient The setting rule is: when the safety margin When the value exceeds a preset threshold, the coefficient remains at a constant base value; when the safety margin... When the value is less than a preset threshold and approaches zero, the coefficient exhibits a non-linear monotonically increasing trend. The control unit utilizes this non-linear gain coefficient to weight and amplify the virtual elastic restoring force. .

[0030] The control unit performs a baseline speed adaptive correction step, sets a sliding time window with a length of 30 to 60 seconds, and calculates the virtual adjustment correction amount within this window. statistical average value , when the absolute value of the difference exceeds a preset deadband threshold, the control unit compensates and corrects the base transport speed based on the statistical average value, for example, using the formula , where is the integral learning rate, and the corrected base transport speed is used as the updated reference value for the generation of subsequent dynamic speed commands. This step also includes state freezing logic, i.e., when it is detected that the production line is in a non-steady state condition such as start-up, emergency stop, or fault alarm, the correction step is suspended, and the current base transport speed is maintained. The control unit performs a resonance active suppression step, in which the control unit monitors the sign flipping frequency of the virtual adjustment correction quantity in the time domain to estimate the real-time oscillation frequency of the collaborative control system. When the real-time oscillation frequency falls within a preset intrinsic resonance frequency band, the control unit applies a time-varying offset to the virtual elastic stiffness coefficient in the virtual viscoelastic dynamic model to change the natural frequency of the collaborative control system. At the same time, the control unit temporarily increases the weight of the virtual viscous damping force. The control unit performs an inertia adaptive gain scheduling step based on the material inventory, which maps the material inventory in real time to an inertia factor representing the current load inertia of the system and dynamically adjusts the virtual elastic stiffness coefficient and the virtual damping coefficient in real time according to a preset nonlinear mapping rule. The nonlinear mapping rule is set as follows: as the inertia factor increases, the virtual elastic stiffness coefficient and the virtual damping coefficient are correspondingly increased to maintain the natural frequency and damping ratio of the collaborative control system constant.

[0031] The control unit vectorially combines the virtual elastic restoring force with the virtual viscous damping force generated based on the asymmetric anisotropic damping generation strategy to generate a virtual adjustment correction quantity , and adds this correction quantity to the base transport speed of the process unit to generate a dynamic speed command for the driving transmission mechanism. Before output, the dynamic speed command is subjected to nonlinear saturation constraint processing: the control unit presets a speed constraint range based on the minimum residence time or maximum physical beat limit of the process step performed by the process unit. When the calculated dynamic speed command exceeds this range, the control unit clamps its value at the boundary value.

[0032] Example 1: In a cascaded high-throughput photovoltaic cell production scenario including screen printing, high-temperature sintering, and automated inspection units, the collaborative control method of this invention is applied to solve the nonlinear cascaded oscillation problem caused by process cycle mismatch. The production line is designed with a single-line capacity of 7200 cells / hour. Adjacent process units are coupled through a physical buffer zone with a capacity of 500 cells. Under continuous operation, due to the reciprocating periodic characteristics of the upstream screen printing machine's squeegee movement, its output cycle exhibits periodic fluctuations with a frequency of approximately 0.5Hz, and the amplitude of this fluctuation reaches ±1 of the average cycle time. 5%, while the downstream detection unit is limited by the processing latency of the visual algorithm, and the receiving cycle has a random jitter of about 200ms. This dynamic mismatch between upstream and downstream cycles, under the traditional discrete threshold control mode, can easily cause the material inventory in the physical buffer to switch drastically between full load and empty state in a short period of time, causing frequent emergency stops and full-speed starts of the transmission mechanism, which in turn causes the microcrack rate of ultrathin silicon wafers (thickness less than 110μm) to rise to more than 0.5%. When the system faces the above conditions, the control unit of this invention collects the current material inventory in the associated physical buffer in real time. and the rate of change of material inventory The control unit uses the constructed virtual viscoelastic dynamics model to calculate the virtual adjustment correction amount in real time. In response to the periodic discharge surges generated by the upstream screen printing machine, namely When the system exhibits large positive fluctuations, its inherent asymmetric anisotropic damping generation strategy is triggered, and the control unit recognizes this. The accumulation trend automatically selects the first damping coefficient with a larger value. This overdamping characteristic immediately generates a strong virtual viscous damping force, simulating a high-viscosity fluid's stagnation effect at the control algorithm level. This suppresses the rapid rise in material inventory and prevents physical stacking caused by upstream surges directly transmitting to the downstream. When the upstream scraper returns, causing a temporary interruption in material discharge, i.e. When the voltage turns negative, the control unit quickly switches to the smaller second damping coefficient. This underdamped characteristic allows the system to be subjected to only weak damping, enabling a virtual elastic stiffness coefficient. The dominant elastic recovery force quickly drives the transmission mechanism to accelerate, and uses the material accumulated in the buffer area to quickly fill the downstream cycle gap, avoiding the risk of downstream equipment starvation shutdown.

[0033] In this process, the system not only achieves flexible adjustment at the single-machine level, but also realizes the coordination of multi-level units through a virtual backpressure communication mechanism. When the downstream detection unit is temporarily blocked due to re-inspection, the virtual pressure value generated by the control unit... Rapidly rising, and passing through the industrial bus to the upstream sintering furnace control unit, after the upstream control unit receives the back pressure signal, as a feedforward inhibition variable, it is directly superimposed into the speed command generation logic, before the physical buffer zone reaches the full load threshold, it performs a predictive deceleration, this control strategy based on information flow advance transmission, converts the hard blockage that may form at the downstream detection unit into a small speed attenuation distributed along the entire production line, avoiding local avalanche effect, in addition, for the inherent 0.5Hz periodic disturbance of the screen printer, the system's resonance active suppression step monitors the virtual adjustment correction amount The sign flip frequency of which falls into the intrinsic resonance frequency band, then a time-varying offset is applied to the virtual elastic stiffness coefficient This active detuning operation destroys the resonance phase condition, so that the material flow in the physical buffer zone maintains a smooth laminar flow state throughout the production cycle, without appearing standing wave oscillation, finally, while maintaining a high throughput of 7200 wafers / hour, the wafer crack rate is reduced to below 0.05%, and the overall equipment effectiveness (OEE) is improved by 3.5%, this result confirms that through the construction of a virtual physical field, the invention successfully converts the rigid physical constraints in discrete manufacturing systems into flexible logical connections with adaptive ability, solving the industry problem of balancing efficiency and stability in high-throughput photovoltaic production lines.

[0034] Example 2: To verify the actual engineering effectiveness and performance superiority of the cooperative control method of the invention under complex working conditions, a semi-physical simulation test platform is constructed, which includes three serial process units (upstream printer, middle stream sintering furnace, downstream detector), relying on the EtherCAT real-time industrial bus architecture, the controllers of each process unit use Beckhoff CX2040 high-performance embedded PC, the control period is set to 2ms, the test simulates the complex disturbance environment in real photovoltaic production: the discharge rhythm of the upstream printer is set to an average of 7200 wafers / hour, and a sine wave with an amplitude of ±15% and a frequency of 0.5Hz is superimposed to simulate the periodic surge caused by the reciprocating motion of the doctor blade; the input rhythm of the downstream detector superimposes a Gaussian white noise with a mean of 0 and a variance of 200ms to simulate the randomness of the vision detection processing delay, in addition, to reflect the electromagnetic interference in the industrial field, a wideband noise with a signal-to-noise ratio of 25dB is mixed into the collected material inventory signal.

[0035] The test design three group schemes to carry out multidimensional comparison and verification, the control group A (prior art group) adopts the traditional high and low liquid level threshold switch control strategy, sets the high threshold value as 80%, and the low threshold value as 20%, the control group B (partially missing type control group) adopts the virtual viscoelastic dynamics model of the application, but removes the asymmetric anisotropic damping generation strategy and the virtual back pressure communication mechanism, only retains the linear virtual impedance control, the test group (the sample group of the application) completely enables all core control logics of the application, including the asymmetric damping and the virtual back pressure mechanism, each group is continuously operated for 4 hours under the same input working condition, and the inventory fluctuation range of the physical buffer zone, the speed fluctuation rate (speed standard deviation / average speed) of the transmission mechanism and the acceleration peak value are monitored; the material inventory and the change rate after 5Hz low pass filtering are collected in real time during the test and the change rate in the test group, when the upstream periodic surge is monitored to cause , the system automatically calls the strong damping coefficient , and suppresses the inventory overshoot through the over-damping characteristic; when , the weak damping coefficient is switched to, and the under-damping characteristic is used to quickly replenish the material, in addition, when the downstream detection machine is simulated to have a continuous 10-second random blockage, the test group uses the virtual back pressure mechanism to make the upstream sintering furnace receive signal and perform pre-deceleration in advance when the local buffer is not full, and table 1 shows the comparison data table of the key performance indicators of the three group schemes.

[0036] Table 1: Comparison data table

[0037]

[0038] The data shows that, due to the discontinuity of the control mechanism, the system of the control group A falls into a severe start-stop limit cycle oscillation, resulting in an acceleration peak value of 5.2g, far exceeding the safety threshold of the silicon wafer, although the virtual impedance is introduced in the control group B to realize continuous adjustment, when responding to the asymmetric material discharge fluctuation and the downstream sudden blockage, due to the lack of directional damping difference and feed-forward back pressure information, the response is still lagged, and the speed fluctuation rate is 12.3%, in comparison, the test group stabilizes the inventory fluctuation range in a narrow band interval of 42% to 58% by virtue of the directional suppression of the asymmetric damping to the surge and the advanced prediction of the back pressure mechanism to the blockage, and the speed fluctuation rate is only 4.8%, and the acceleration peak value is suppressed within the safety range of 1.1g, which confirms that the application fundamentally reconstructs the dynamics characteristics of the production line logistics through the deep cooperation of the virtual physical field, the asymmetric damping and the back pressure communication, and realizes high stability and low damage transmission under complex disturbance.

[0039] Embodiment 3: This embodiment combines Figs. 1 to 3The collaborative control method for photovoltaic intelligent manufacturing equipment production lines is explained, such as... Fig. 1 As shown, the collaborative control method performs a physical buffer state acquisition step to monitor the material inventory and change rate in real time. The process is divided into parallel processing logics, which respectively execute an asymmetric anisotropic damping generation strategy that dynamically switches the damping coefficient according to the accumulation or evacuation trend, establish a virtual viscoelastic dynamic model that maps non-Newtonian fluid characteristics, and introduce a virtual back pressure communication mechanism to suppress downstream blockage pressure feedforward. The calculation results of the above branches are incorporated into the virtual adjustment correction synthesis step to complete the vector superposition of elastic restoring force and viscous damping force, and then enter the reference speed adaptive correction step to eliminate the steady-state deviation of the system. Finally, in the dynamic speed command generation step, the transmission mechanism is driven after nonlinear saturation constraint processing.

[0040] like Fig. 2 As shown, the horizontal axis represents time (seconds), the left vertical axis represents the dimensionless virtual force, and the right vertical axis represents the material inventory change rate Q (% / second). The graph shows that the material inventory change rate Q exhibits periodic sinusoidal fluctuations, and the associated virtual viscous damping force curve shows asymmetric anisotropy characteristics, that is, a large-amplitude inhibiting force is generated in the range where Q is positive, while a small-amplitude damping force is generated in the range where Q is negative; for example... Fig. 3 As shown, the architecture of the collaborative control system is centered on the process unit control unit. The system collects the inventory and rate of change signals of the material state in real time through sensor components. The control unit integrates a virtual viscoelastic dynamic model calculation module constrained by a nonlinear boundary hardening strategy, an asymmetric anisotropic damping generation strategy module, a resonance active suppression module responsible for frequency monitoring and detuning, and a reference speed adaptive correction module. At the same time, the control unit receives virtual back pressure communication feedforward suppression variables from downstream process units. All calculation results are summarized to the dynamic speed command generation module, and after vector synthesis and saturation constraint, the final control command is output to the transmission mechanism.

[0041] Example 4: In the development of a highly integrated photovoltaic intelligent manufacturing equipment production line control system, addressing the engineering calibration challenges of parameter setting in the virtual viscoelastic dynamics model, this invention proposes and implements a systematic parameter optimization and calibration procedure. This procedure aims to solve the problem of virtual elastic stiffness coefficient... With virtual damping coefficient In practical applications, parameter issues arise due to the lack of explicit physical mapping. To ensure the stability and response performance of the control model under different operating conditions, it is crucial to clarify the calibration object and influencing factors. The calibration object is the core parameters in the virtual viscoelastic dynamics model. and , the main technical factors affecting the parameter values include: the effective capacity of the physical buffer zone, the maximum acceleration and deceleration capability of the transmission mechanism, the fluctuation frequency and amplitude of the upstream outfeed rhythm, and the sensitivity of the downstream process unit to the continuous supply of materials. The essence of technical trade-off lies in: higher can improve the response speed of the system to inventory deviation, but too high can cause system overshoot and even oscillation; higher can enhance the stability and anti-disturbance ability of the system, but too high can increase the viscosity of the system, causing response delay, based on the above analysis, the decision logic and operation process of parameter calibration are established, first, perform offline basic parameter calibration, in the empty state of the production line, set the physical buffer zone to half full state, that is , and apply a stepwise speed disturbance signal, adjust until the system's rise time meets the shortest requirement of the process rhythm and the overshoot is less than 5%, at this time, the value recorded as the basic stiffness coefficient, keep unchanged, gradually increase , until the system's oscillation decay ratio reaches decay standard, that is, the second peak amplitude is 25% of the first peak amplitude, the value at this time as the basic damping coefficient.

[0042] Second, perform online working condition adaptive calibration, during the running process of the production line under load, enable the inertia adaptive gain scheduling mechanism, the system real-time collects the material inventory , and maps it to the dimensionless inertia factor , according to the principle of dynamic similarity, set in direct proportion to , that is , to maintain the inherent frequency of the system approximately constant under different loads; at the same time, set in direct proportion to , that is , to maintain the damping ratio of the system constant, this mapping rule ensures that the system always has consistent dynamic response characteristics in the full working condition range from empty load to full load; third, perform fine tuning of asymmetric damping characteristics, for the asymmetric anisotropic damping generation strategy, set the first damping coefficient (corresponding to the accumulation trend) and the second damping coefficient (corresponding to the dispersion trend), in actual production, when the accumulation trend of materials is monitored , set as 1.5 to 2.0 times the basic damping coefficient, to build strong damping characteristics, to preferentially suppress overshoot and overflow risk; when the dispersion trend of materials is monitored When ), The damping coefficient is set to 0.5 to 0.8 times the base damping coefficient to construct weak damping characteristics. The elastic restoring force is used to accelerate the feeding. Through long-term operation tests, the above ratio coefficient is fine-tuned until the comprehensive evaluation indicators of the system, including OEE, microcrack rate, and mean time between failures, reach the optimal level.

[0043] Example 5: To ensure the stability and control accuracy of the asymmetric damping generation strategy and virtual backpressure communication mechanism of this invention in actual industrial deployment, this invention constructs a standardized on-site pre-deployment calibration and debugging procedure. This aims to eliminate control model mismatches that may be caused by individual equipment differences, sensor installation errors, and fluctuations in the on-site electromagnetic environment through systematic parameter identification and dynamic optimization. Technicians need to perform on-site calibration of the effective capacity of the physical buffer zone. Under no-load conditions, the control unit drives the upstream equipment to output material in a step manner. Photoelectric sensors installed at the inlet and outlet of the buffer zone record the time difference of material passage, and the actual maximum capacity of the buffer zone is calculated and confirmed. This calibration value is written into the control algorithm as a benchmark for material inventory normalization calculation to ensure state deviation. The calculation accuracy is good for asymmetric damping coefficients. and The system was configured to perform dynamic response testing. After connecting a simulated load to the system, a sinusoidal disturbance signal with a frequency of 0.1Hz to 1.0Hz was artificially introduced as a virtual beat input. By monitoring the speed response curve of the transmission mechanism, adjustments were gradually made. Until the system's overshoot to positive disturbances is less than 5%.

[0044] The damping value recorded at this point serves as the basic damping coefficient under the accumulation trend; similarly, adjust... Until the system's response time to negative disturbances meets the minimum cycle time requirement of downstream equipment, the basic damping coefficient under the evacuation trend is determined. This debugging step ensures that the asymmetric damping mechanism can accurately match the physical inertia characteristics of the current production line. Finally, a stress test of the virtual backpressure communication link is conducted. Under full-load operation, a continuous blockage of the downstream unit is simulated, and the virtual pressure value received by the upstream unit is monitored. The trend of change and response time; if the response delay exceeds a preset threshold, such as 500ms, the pressure conduction attenuation coefficient needs to be adjusted. To optimize the reverse transmission efficiency of pressure signals and ensure the real-time performance of the entire line's coordinated control, the system completes the adaptation from the theoretical model to the engineering site through the execution of the above procedures, ensuring the reliability of the coordinated control strategy.

[0045] Example 6: To verify the stability of the synergistic control method of the application under extreme working conditions and long-term operation, an offline physical parameter identification sub-procedure is constructed for the geometry and inertia characteristics of the physical buffer zone. When the production line is in a shutdown maintenance state, a laser range finder is used to measure the effective transmission length of the physical buffer zone between each process unit Precise measurements are taken, the drive transmission mechanism is operated at step speed in the empty state, the speed response data is collected by a high-frequency encoder, and the least squares method is used to identify the equivalent rotational inertia of the system And the base friction coefficient These physical parameters are written into the underlying configuration file of the controller as core constants, providing a definite physical benchmark for the construction of the subsequent virtual dynamics model. For the setting of key control parameters in the asymmetric damping generation strategy, an online dynamic response optimization sub-procedure is executed. During the system trial operation phase, a variable frequency sinusoidal speed disturbance signal with a frequency of 0.1 Hz to 2.0 Hz is injected into the control loop, and the control unit monitors the speed tracking error of the transmission mechanism in real time Through an adaptive optimization algorithm, the first damping coefficient And the second damping coefficient are adjusted respectively until the overshoot under material accumulation (positive disturbance) conditions is less than 3%, and the regulation time under material dispersion (negative disturbance) conditions is less than 20% of the minimum beat of the downstream equipment. At this time, the determined And are fixed as the optimal damping parameter set for the current working condition of the production line.

[0046] Again, for the real-time and stability of the virtual back pressure communication mechanism, a network delay and packet loss stress test sub-procedure is executed. Using a network analyzer to simulate the industrial bus communication environment under high load, random packet loss and time delay jitter are injected into the control network, and the integrity and real-time of the virtual pressure value received by the upstream control unit are monitored. If there are more than 3 packet losses or a single time delay exceeds 10 ms within 1 minute, the system automatically triggers the degradation protection logic, temporarily switches to the local independent control mode, and records the abnormal log for subsequent network optimization reference, ensuring the safe failure of the synergistic control when the communication link is unstable. Finally, for the setting of the trigger threshold of the nonlinear boundary hardening strategy, a boundary safety margin verification sub-procedure is executed. Under full load operation, a completely blocked downstream condition is artificially created, and the material inventory is forced to approach the physical limit. The change curve of the nonlinear gain coefficient with the safety margin is recorded in real time, and the curvature parameter of the nonlinear function is adjusted until the virtual elastic restoring force can generate a reverse braking acceleration sufficient to overcome the maximum static friction of the system before the material inventory reaches 5% of the physical limit, thereby achieving soft landing protection without the intervention of mechanical limit switches.

[0047] Example 7: Calibration of key control parameters for virtual viscoelasticity dynamics model Initial value is determined by applying a step velocity command to the empty load transfer mechanism and monitoring the material inventory Response curve determination, adjust the coefficient to match the response rise time to the minimum physical beat of the downstream process unit and the overshoot to be within the preset range; keep Stepwise increase the basic damping coefficient to the inventory response curve oscillation decay ratio to meet the quarter decay standard, based on the basic damping value, the control unit adjusts the first damping coefficient according to the material inventory change rate Sign direction, the first damping coefficient for the accumulation trend Set to 1.5 to 2.0 times the basic value to strengthen overshoot suppression, the second damping coefficient for the dispersion trend Set to 0.5 to 0.8 times the basic value to improve the follow-up response speed.

[0048] Resonance active suppression step oscillation frequency extraction uses hysteresis interval zero-crossing detection logic to process virtual adjustment correction Time domain signal, set the hysteresis interval threshold to 5% to 10% of the signal full scale, and record it as an effective zero-crossing event when the signal amplitude crosses the interval and the polarity flips, shielding the zero point jitter of the field electromagnetic interference or mechanical microseismic signal; the control unit calculates the statistical frequency of the effective zero-crossing events in the preset sliding time window, and when the statistical frequency falls within the 0.8 to 1.2 times frequency band range of the system first-order natural frequency determined by modal analysis or sweep test, the virtual elastic stiffness coefficient Time-varying offset modulation, change the equivalent stiffness of the control loop to make the system closed-loop pole deviate from the mechanical resonance point; virtual back pressure communication mechanism generates a virtual pressure value According to the current material inventory Relative to the limit capacity of the physical buffer area Local pressure component is calculated, and the local pressure component is superimposed with the received downstream virtual pressure value Weighted algorithm pressure conduction attenuation coefficient Depends on the cascade length of the production line physical layout, the direct coupling coefficient between adjacent nodes is set to 0.8 to 0.95, ensuring that the congestion information is transmitted to at least three upstream process units in the counter-flow direction, and using the step-by-step attenuation characteristic to prevent small disturbances in the long-chain transmission from being amplified too much to trigger a full-line speed reduction response.

[0049] It is apparent to those skilled in the art that the present application is not limited to the details of the above exemplary embodiments, and can be implemented in other specific forms without departing from the spirit or essential characteristics of the present application.

[0050] Finally, it should be noted that the above examples are merely intended to illustrate the technical solutions of the present application and not to limit the present application. Although the present application is described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present application.

Claims

1.A method for collaborative control of a photovoltaic smart manufacturing equipment line, characterized in that, The production line includes a plurality of process units connected in series through physical buffer zones, and the method maps discrete material flows in the physical buffer zones to a virtual viscoelastic dynamic model with non-Newtonian fluid characteristics, and the method comprises the following steps: The control unit of each process unit collects the current material inventory in the associated physical buffer zone and the change rate of the material inventory in real time; The control unit calculates a virtual elastic restoring force that returns the material inventory to a preset equilibrium point and a virtual viscous damping force that hinders the change of the material inventory state based on the virtual viscoelastic dynamic model; Wherein, in the calculation of the virtual viscous damping force, the control unit performs an asymmetric anisotropic damping generation strategy: it identifies the positive and negative polarity of the change rate of the material inventory in real time to determine the instantaneous trend of the material flow, when it is determined that the trend is material accumulation, a first damping coefficient is called to calculate an over-damping force with strong inhibition characteristics, and when it is determined that the trend is material dispersion, a second damping coefficient smaller than the first damping coefficient is called to calculate an under-damping force with weak inhibition characteristics; The control unit vectorially combines the virtual elastic restoring force and the virtual viscous damping force obtained based on the asymmetric anisotropic damping generation strategy to generate a virtual adjustment correction, and superimposes the virtual adjustment correction on the basic transmission speed of the process unit to generate a dynamic speed instruction for driving the transmission mechanism; Through the asymmetric anisotropic damping generation strategy, the process unit exhibits a nonlinear rheological characteristic that the response to upstream flow fluctuations coexists with rigid inhibition in the accumulation trend and flexible following in the dispersion trend; In the virtual viscoelastic dynamic model, the calculation logic of the virtual adjustment correction amount follows a linear combination relationship containing a state deviation term and a state change rate term, and satisfies the following dynamic equation: wherein, characterizing a virtual adjustment correction amount, characterizing a virtual elastic stiffness coefficient, characterizing a current material inventory characterizing a preset equilibrium point, characterizing a flow trend a first damping coefficient or a second damping coefficient switched dynamically, characterizing a change rate of the material inventory; the equation defines the dynamic convergence trajectory of the system when deviating from the steady state. 2.The photovoltaic intelligent manufacturing equipment line collaborative control method according to claim 1, characterized in that, The method further comprises establishing a virtual back pressure communication mechanism that transmits in the reverse direction along the material flow direction of the production line: the control unit receives the virtual pressure value sent by the directly downstream process unit, and introduces the virtual pressure value as a feedforward inhibition variable into the generation logic of the dynamic speed instruction; the virtual pressure value is generated by the downstream process unit based on its own current material inventory and the more downstream virtual pressure value it receives through weighted accumulation calculation; through the virtual back pressure communication mechanism, the material flow blockage state of the downstream is diffused in the form of a scalar field to the upstream in the reverse direction, driving the upstream process unit to perform predictive deceleration before the physical buffer zone is full, and using the lead transmission of information flow to exchange for the redundancy of physical buffer space. 3.The photovoltaic intelligent manufacturing equipment line collaborative control method according to claim 1, characterized in that, The step of calculating the virtual elastic restoring force includes performing a nonlinear boundary hardening strategy: the control unit monitors the safety margin between the material inventory and the physical limit boundary of the physical buffer zone in real time; the control unit generates a nonlinear gain coefficient based on the safety margin, which is set to remain a constant base value when the safety margin is greater than a preset threshold, and is nonlinearly and monotonously increasing when the safety margin is less than the preset threshold and approaches zero; the control unit uses the nonlinear gain coefficient to weight and amplify the virtual elastic restoring force to generate an exponentially enhanced repulsive potential field when the material inventory approaches the physical limit boundary, so that the transmission mechanism continuously corrects the speed in the opposite direction without external hard limit triggering. 4.The photovoltaic intelligent manufacturing equipment line collaborative control method of claim 1, wherein, The method further comprises a reference speed adaptive correction step: the control unit sets a sliding time window, and calculates a statistical average of the virtual adjustment correction amount within the window; when the absolute value of the statistical average exceeds a preset dead zone threshold, the control unit identifies that there is a systematic speed mismatch, and compensates and corrects the basic transmission speed based on the statistical average, and uses the corrected basic transmission speed as an updated reference value for generation of a subsequent dynamic speed instruction; Through the correction step, the steady-state inventory deviation caused by equipment aging or environmental drift is eliminated, the virtual adjustment correction amount automatically converges to the vicinity of zero under steady-state operation, and the dynamic range of the system bidirectional adjustment is restored. 5.The photovoltaic intelligent manufacturing equipment line collaborative control method according to claim 1, wherein, The method further comprises a resonance active suppression step: the control unit monitors the sign flip frequency of the virtual adjustment correction amount in the time domain to estimate the real-time oscillation frequency of the cooperative control system; when the real-time oscillation frequency falls within a preset intrinsic resonance frequency band, the control unit applies a time-varying offset to the virtual elastic stiffness coefficient in the virtual viscoelastic dynamic model, dynamically changes the natural frequency of the cooperative control system to destroy the resonance phase synchronization condition; the control unit temporarily increases the weight of the virtual viscous damping force when the resonance trend is detected to dissipate the resonance energy in the system. 6.The photovoltaic intelligent manufacturing equipment line collaborative control method according to claim 1, characterized in that, The method further comprises an inertia adaptive gain scheduling step based on the material inventory: the control unit maps the material inventory in real time to an inertia factor representing the current load inertia of the system; the control unit dynamically adjusts the virtual elastic stiffness coefficient and the virtual damping coefficient in the virtual viscoelastic dynamic model based on the inertia factor and according to a preset nonlinear mapping rule; The nonlinear mapping rule is set to correspondingly increase the virtual elastic stiffness coefficient and the virtual damping coefficient as the inertia factor increases, so as to maintain the natural frequency and damping ratio of the cooperative control system constant under different load conditions, and ensure the consistency of the system dynamic response characteristics. 7.The photovoltaic intelligent manufacturing equipment line collaborative control method according to claim 1, wherein, The acquisition of the change rate of the material inventory in the state acquisition step includes: acquiring the material inventory difference value of the adjacent two sampling periods; performing low-pass filtering processing on the material inventory difference value to filter out high-frequency noise interference, and generating a smoothed change rate signal as a basis for calculating the virtual viscous damping force; the cutoff frequency of the low-pass filtering processing is set to be lower than the lowest effective response frequency of the cooperative control system, to prevent the asymmetric anisotropic damping generation strategy from being triggered by sensor signal jitter. 8.The photovoltaic intelligent manufacturing equipment line collaborative control method according to claim 1, wherein, The step of generating a dynamic speed instruction further comprises nonlinear saturation constraint processing: the control unit presets a speed constraint range based on the minimum residence time or maximum physical beat limit of the process step performed by the process unit; when the calculated dynamic speed instruction exceeds the speed constraint range, the control unit clamps the dynamic speed instruction at the boundary value of the speed constraint range, to preferentially guarantee that the single-machine process parameters do not exceed the limit. 9.The photovoltaic intelligent manufacturing equipment line collaborative control method of claim 4, wherein, The reference speed adaptive correction step further comprises a state freezing logic: the control unit monitors the running state of the production line in real time; when it is detected that the production line is in a non-steady-state condition such as start-up, emergency stop or fault alarm, the control unit automatically suspends the reference speed adaptive correction step, and keeps the current basic transmission speed unchanged.

Images