A decentralized control system for a rapeseed oil pressing process

Abstract

This invention belongs to the field of industrial automation and program control technology, and discloses a distributed control system for rapeseed oil pressing process, comprising: a steaming and frying control node, a feeding adjustment control node, a pressing main drive control node, a cake discharge adjustment control node, and a deterministic bus network. The pressing main drive control node constructs a sliding sampling window to calculate the wall temperature change rate, calculates the pressing stress state parameters in combination with the main drive motor operating current, and broadcasts differential state data frames without a master when the parameters exceed the threshold. The feeding adjustment control node and the cake discharge adjustment control node receive the data frames, respectively reduce the operating frequency of the feeding motor and expand the cake discharge gap. This invention eliminates the scheduling delay of the central node and bus queuing jitter, predicts the stress change of the pressing chamber in advance, avoids jamming faults, and reduces the residual oil rate of the oil cake.

Description

Technical Field

[0001] This invention relates to a distributed control system for rapeseed oil pressing, belonging to the field of industrial automation and program control technology. Background Technology

[0002] In the continuous pressing stage of an industrial automated production line, a distributed control system is typically used, distributing different control tasks to various control nodes to coordinate processes such as steaming and roasting pretreatment, screw pressing, and cake discharge. The moisture and oil content of rapeseed raw materials naturally fluctuate, causing frequent and significant changes in the load on the main drive shaft of the screw press. In a centralized master-slave control architecture, a central master node is required to aggregate data and issue unified control commands. Network transmission and data processing take time, easily causing the control response to lag behind the load changes inside the pressing chamber, thus triggering the overcurrent protection of the main drive motor or causing the screw press to jam. When each node independently performs single-loop adjustment, the lack of coordination of status information between nodes makes it difficult for the feeding node to make timely feedforward adjustments based on the stress changes inside the pressing chamber, easily causing system oscillations and resulting in residual oil in the cake. In addition to the limitations of mechanical structure and conventional single-loop control, the existing automated control methods based on high-order algorithms also have shortcomings in dealing with the physical abrupt changes in the continuous pressing process when the oil yield increases. For example, Chinese invention patent application CN114661008A discloses an oilseed processing control system and method based on cloud-edge collaboration. It uses cloud neural network modeling and edge collaboration to regulate the production line. However, the abrupt changes in material friction and extrusion stress during continuous pressing are millisecond-level nonlinear physical processes. The cloud-edge collaboration architecture relies on remote IP network transmission and complex model inference. There are time delays that are difficult to eliminate in the data interaction and calculation links, which causes the control response to lag behind the stress evolution inside the pressing chamber. It is impossible to complete the feedforward adjustment before jamming occurs, and it is still difficult to avoid the risks of overload shutdown and product quality fluctuations.

[0003] The screw pressing process involves strong mechanical vibrations and high temperatures in the pressing area. High-frequency electromagnetic interference and random thermal noise easily infiltrate the signals from the main drive motor's operating current and the pressing chamber wall temperature. When calculating the rate of change of wall temperature and the stress increment over time, if discrete differential calculation is used directly, the noise peaks in the signal are amplified during the differential process, easily causing control malfunctions. Using traditional low-pass filtering introduces phase lag, preventing timely reflection of actual stress abrupt changes. Furthermore, simply increasing network bandwidth or improving the processing performance of the central node cannot eliminate the transmission delay caused by bus arbitration and data queuing among multiple nodes. Increasing the controller gain may further amplify noise in the sensor signals, exacerbating system oscillations. Existing distributed control systems used in continuous pressing processes still face constraints between cross-node control response, noise amplification in differential calculations, and the communication burden between nodes, making it difficult to balance deterministic real-time communication with adaptive coordination between control nodes.

[0004] Therefore, how to reduce the latency caused by central node scheduling and bus transmission, while timely judging the stress change trend in the pressing chamber and realizing feedforward coordination between control nodes, has become a technical problem that needs to be solved in this field. Summary of the Invention

[0005] To address the problems in the background art, the technical solution of the present invention is as follows: A decentralized control system for rapeseed oil pressing, comprising:

[0006] The steaming and frying control node includes a steaming and frying temperature acquisition module and an output humidity acquisition module, which acquires the rapeseed steaming and frying temperature and output humidity, and outputs steam valve control signals for steaming and frying.

[0007] The feeding adjustment and control node includes a feeding drive module and a speed monitoring module, which collects the operating frequency of the feeding motor and the speed of the feeding mechanism, and outputs feeding drive commands.

[0008] The main drive control node for pressing includes a main drive current acquisition module, a wall temperature acquisition module, and a stress parameter calculation module. It acquires the operating current of the main drive motor and the wall temperature of the pressing chamber, constructs a 4-point sliding sampling timing window with a communication cycle locked with the deterministic bus network, calculates the denoised wall temperature change rate using the wall temperature sequence of 4 consecutive sampling times, and calculates the pressing stress state parameters. When the pressing stress state parameters exceed the preset stress disturbance threshold, it generates a differential state data frame containing stress increment and timestamp, and broadcasts it to the feed adjustment control node and the cake discharge adjustment control node without a master through the deterministic bus network.

[0009] The feed adjustment control node receives differential state data frames, calculates the rate of change of stress increment over time using a deterministic time-based second-order backward differential operator based on the timestamp, and lowers the operating frequency of the feed motor.

[0010] The cake discharge adjustment control node includes a cake discharge gap adjustment module, which receives differential state data frames and controls the cake discharge gap adjustment module to expand the cake discharge gap according to the stress increment.

[0011] Preferably, the deterministic bus network adopts a time-triggered deterministic bus scheduling protocol; the deterministic bus network opens a periodic cyclic sequence of time slots within the communication cycle, and divides the time slot sequence into deterministic real-time transmission time slots and non-deterministic non-real-time transmission time slots; the master drive control node is forced to directly map the differential state data frame into the reserved deterministic real-time transmission time slot for masterless broadcasting, avoiding arbitration and queue buffering jitter when non-real-time data generated by other discrete nodes are transmitted on the bus, and controlling the transmission delay of the differential state data frame on the bus within the preset upper limit of the delay.

[0012] Preferably, the feed adjustment control node calculates the reduction amount of the feed motor operating frequency by weighted summing of the extracted stress increment and its rate of change over time based on the preset gain ratio coefficient and differential adjustment coefficient. When the reduction amount of the feed motor operating frequency is greater than the preset single-step maximum frequency reduction limit, the feed adjustment control node cuts off and locks the reduction amount of the feed motor operating frequency to the single-step maximum frequency reduction limit, and outputs a frequency reduction command to the feed drive module to smooth out the mechanical impact caused by the deceleration of the feed mechanism. After the rate of change of the stress increment over time returns to zero or a negative value, the feed adjustment control node gradually restores the feed motor operating frequency according to the preset step size.

[0013] Preferably, the steaming and frying control node is connected to a deterministic bus network and listens to differential state data frames; when a differential state data frame indicating that the stress increment is greater than zero is continuously received within a preset observation time window, the steaming and frying control node calculates the cumulative integral value of the stress increment within the observation time window; the steaming and frying control node calculates the amount of adjustment of the steaming and frying steam valve opening based on the product of the cumulative integral value and the preset steaming and frying integral compensation coefficient, and adjusts the opening of the steaming and frying steam valve.

[0014] Preferably, when the cake dispensing adjustment control node receives the differential state data frame, it extracts the cake dispensing pressure at the dispensing port and the position of the cake dispensing mechanism. The cake dispensing adjustment control node multiplies the stress increment by a preset cake dispensing gap adjustment gain to calculate the cake dispensing gap expansion amount. When the cake dispensing pressure at the dispensing port exceeds a preset safe pressure threshold, the cake dispensing adjustment control node controls the cake dispensing gap adjustment module to expand the cake dispensing gap by the cake dispensing gap expansion amount. When the position of the cake dispensing mechanism reaches a preset maximum safe travel limit, the cake dispensing adjustment control node stops expanding the cake dispensing gap and sends a cake dispensing limit alarm frame to the deterministic bus network.

[0015] Preferably, when constructing a 4-point sliding sampling time window, the main drive control node of the pressing chamber stores the wall temperature of the pressing chamber collected in 4 consecutive sampling cycles in a ring-shaped sliding buffer. The main drive control node of the pressing chamber assigns discrete weighting coefficients that decay with time to the wall temperature at the current sampling time, the wall temperature at the previous sampling time, the wall temperature at the second sampling time, and the wall temperature at the third sampling time, calculates the weighted discrete gradient, and extracts the wall temperature change rate.

[0016] Preferably, when the pressing process includes multiple screw press units configured in parallel, each screw press unit is connected to a pressing main drive control node and a feed adjustment control node respectively; each pressing main drive control node shares its own main drive motor operating current derivative in real time through a deterministic bus network in a peer-to-peer communication manner; when the operating current derivative of a certain screw press unit is continuously higher than the average current derivative of the other screw press units, the corresponding pressing main drive control node generates a current shunting adjustment data frame, outputs a current shunting adjustment command to the feed adjustment module, reduces the current shunted to that screw press unit, and increases the current shunted to the other screw press units whose operating current derivatives are within the normal range.

[0017] Preferably, the distributed control system further includes an oil outlet self-cleaning node; the oil outlet self-cleaning node collects the pressure difference on both sides of the oil collection tank filter screen of the oil outlet pan and the flow rate of the oil outlet pipeline, and calculates the transient rate of change of the flow rate and pressure difference of the oil outlet pipeline; when the rate of increase of the pressure difference and the rate of decrease of the flow rate both exceed their respective preset thresholds, the oil outlet self-cleaning node controls the slag discharge cleaning module to adjust the action frequency and flushing cycle.

[0018] Preferably, when the pressing stress state parameter calculated by the pressing main drive control node falls below the stress disturbance threshold, the pressing main drive control node stops broadcasting differential state data frames to the deterministic bus network; when the feeding adjustment control node, the cake discharge adjustment control node, and the steaming and frying control node do not receive differential state data frames within the preset timeout period, the feeding motor operating frequency, the cake discharge gap, and the steaming and frying valve opening are restored to the preset closed-loop control baseline according to the preset linear slope.

[0019] Compared with the prior art, the beneficial effects of the present invention are:

[0020] 1. In the decentralized control of rapeseed oil pressing process, the feed rate and cake discharge resistance can be adjusted in a timely manner according to the changes in pressing load by the coordinated linkage of load current and temperature signal. The feeding speed is constrained by temperature change, so that the internal extrusion pressure of the pressing chamber is matched with the thermoplastic state of the material. This can reduce the blockage of the pressing chamber caused by material accumulation and the oil coking caused by local overheating, reduce the operation stagnation and quality fluctuation caused by material conveying delay and pressure change, and make the pressing process more stable.

[0021] 2. By interacting with the control nodes of the steaming and roasting section and the pressing section, the moisture content of the steamed and roasted material is coordinated with the torque of the main shaft of the press. The steaming and roasting and feeding are adjusted in advance according to the pressing status. This can reduce the control lag caused by the material being transported from the steaming and roasting section to the pressing section, so that the rapeseed embryos are kept in a more suitable plasticized state when they enter the pressing process. This reduces the material jamming in the pressing chamber and the incomplete separation of residue and oil caused by the fluctuation of moisture content, and improves the control stability and continuity of the continuous pressing process.

[0022] 3. In multiple parallel oil pressing units, through peer-to-peer communication and flow distribution adjustment between control nodes, the feed flow of a certain unit can be reduced in a timely manner when the load of a certain unit increases, and the material can be distributed to other units with normal loads. This can reduce the impact of communication delay on overload handling during centralized scheduling, make the load distribution between units more balanced, the adjustment process more stable, and reduce the risk of shutdown failure and material jamming caused by sudden overload of a single machine. Attached Figure Description

[0023] Figure 1 This is a block diagram of the distributed control system of the present invention;

[0024] Figure 2 This is a collaborative logic diagram of the distributed control system of the present invention.

[0025] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0026] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0027] A distributed control system for rapeseed oil pressing, comprising:

[0028] The steaming and frying control node includes a steaming and frying temperature acquisition module and an output humidity acquisition module, which acquires the rapeseed steaming and frying temperature and output humidity, and outputs steam valve control signals for steaming and frying.

[0029] The feeding adjustment and control node includes a feeding drive module and a speed monitoring module, which collects the operating frequency of the feeding motor and the speed of the feeding mechanism, and outputs feeding drive commands.

[0030] The main drive control node for pressing includes a main drive current acquisition module, a wall temperature acquisition module, and a stress parameter calculation module. It acquires the operating current of the main drive motor and the wall temperature of the pressing chamber, constructs a 4-point sliding sampling timing window with a communication cycle locked with the deterministic bus network, calculates the denoised wall temperature change rate using the wall temperature sequence of 4 consecutive sampling times, and calculates the pressing stress state parameters. When the pressing stress state parameters exceed the preset stress disturbance threshold, it generates a differential state data frame containing stress increment and timestamp, and broadcasts it to the feed adjustment control node and the cake discharge adjustment control node without a master through the deterministic bus network.

[0031] The feed adjustment control node receives differential state data frames, calculates the rate of change of stress increment over time using a deterministic time-based second-order backward differential operator based on the timestamp, and lowers the operating frequency of the feed motor.

[0032] The cake discharge adjustment control node includes a cake discharge gap adjustment module, which receives differential state data frames and controls the cake discharge gap adjustment module to expand the cake discharge gap according to the stress increment.

[0033] Preferably, the deterministic bus network adopts a time-triggered deterministic bus scheduling protocol; the deterministic bus network opens a periodic cyclic sequence of time slots within the communication cycle, and divides the time slot sequence into deterministic real-time transmission time slots and non-deterministic non-real-time transmission time slots; the master drive control node is forced to directly map the differential state data frame into the reserved deterministic real-time transmission time slot for masterless broadcasting, avoiding arbitration and queue buffering jitter when non-real-time data generated by other discrete nodes are transmitted on the bus, and controlling the transmission delay of the differential state data frame on the bus within the preset upper limit of the delay.

[0034] Preferably, the feed adjustment control node calculates the reduction amount of the feed motor operating frequency by weighted summing of the extracted stress increment and its rate of change over time based on the preset gain ratio coefficient and differential adjustment coefficient. When the reduction amount of the feed motor operating frequency is greater than the preset single-step maximum frequency reduction limit, the feed adjustment control node cuts off and locks the reduction amount of the feed motor operating frequency to the single-step maximum frequency reduction limit, and outputs a frequency reduction command to the feed drive module to smooth out the mechanical impact caused by the deceleration of the feed mechanism. After the rate of change of the stress increment over time returns to zero or a negative value, the feed adjustment control node gradually restores the feed motor operating frequency according to the preset step size.

[0035] Preferably, the steaming and frying control node is connected to a deterministic bus network and listens to differential state data frames; when a differential state data frame indicating that the stress increment is greater than zero is continuously received within a preset observation time window, the steaming and frying control node calculates the cumulative integral value of the stress increment within the observation time window; the steaming and frying control node calculates the amount of adjustment of the steaming and frying steam valve opening based on the product of the cumulative integral value and the preset steaming and frying integral compensation coefficient, and adjusts the opening of the steaming and frying steam valve.

[0036] Preferably, when the cake dispensing adjustment control node receives the differential state data frame, it extracts the cake dispensing pressure at the dispensing port and the position of the cake dispensing mechanism. The cake dispensing adjustment control node multiplies the stress increment by a preset cake dispensing gap adjustment gain to calculate the cake dispensing gap expansion amount. When the cake dispensing pressure at the dispensing port exceeds a preset safe pressure threshold, the cake dispensing adjustment control node controls the cake dispensing gap adjustment module to expand the cake dispensing gap by the cake dispensing gap expansion amount. When the position of the cake dispensing mechanism reaches a preset maximum safe travel limit, the cake dispensing adjustment control node stops expanding the cake dispensing gap and sends a cake dispensing limit alarm frame to the deterministic bus network.

[0037] Preferably, when constructing a 4-point sliding sampling time window, the main drive control node of the pressing chamber stores the wall temperature of the pressing chamber collected in 4 consecutive sampling cycles in a ring-shaped sliding buffer. The main drive control node of the pressing chamber assigns discrete weighting coefficients that decay with time to the wall temperature at the current sampling time, the wall temperature at the previous sampling time, the wall temperature at the second sampling time, and the wall temperature at the third sampling time, calculates the weighted discrete gradient, and extracts the wall temperature change rate.

[0038] Preferably, when the pressing process includes multiple screw press units configured in parallel, each screw press unit is connected to a pressing main drive control node and a feed adjustment control node respectively; each pressing main drive control node shares its own main drive motor operating current derivative in real time through a deterministic bus network in a peer-to-peer communication manner; when the operating current derivative of a certain screw press unit is continuously higher than the average current derivative of the other screw press units, the corresponding pressing main drive control node generates a current shunting adjustment data frame, outputs a current shunting adjustment command to the feed adjustment module, reduces the current shunted to that screw press unit, and increases the current shunted to the other screw press units whose operating current derivatives are within the normal range.

[0039] Preferably, the distributed control system further includes an oil outlet self-cleaning node; the oil outlet self-cleaning node collects the pressure difference on both sides of the oil collection tank filter screen of the oil outlet pan and the flow rate of the oil outlet pipeline, and calculates the transient rate of change of the flow rate and pressure difference of the oil outlet pipeline; when the rate of increase of the pressure difference and the rate of decrease of the flow rate both exceed their respective preset thresholds, the oil outlet self-cleaning node controls the slag discharge cleaning module to adjust the action frequency and flushing cycle.

[0040] Preferably, when the pressing stress state parameter calculated by the pressing main drive control node falls below the stress disturbance threshold, the pressing main drive control node stops broadcasting differential state data frames to the deterministic bus network; when the feeding adjustment control node, the cake discharge adjustment control node, and the steaming and frying control node do not receive differential state data frames within the preset timeout period, the feeding motor operating frequency, the cake discharge gap, and the steaming and frying valve opening are restored to the preset closed-loop control baseline according to the preset linear slope.

[0041] Example 1: The distributed control system in this example is deployed on a rapeseed oil pressing production line that includes steaming and roasting pretreatment, screw pressing, cake discharge and residue separation. It is suitable for operating conditions where the raw material moisture content is high and fluctuates significantly, and the production line experiences continuous and strong mechanical vibration. The steaming and roasting temperature acquisition module and the discharge humidity acquisition module in the steaming and roasting control node acquire the rapeseed steaming and roasting temperature and discharge humidity, respectively, and output the steaming and roasting steam valve control signal. The speed monitoring module in the feeding regulation control node acquires the speed of the feeding mechanism, and the feeding drive module receives the feeding drive command and adjusts the operating frequency of the feeding motor. The pressing main drive control node is connected to a deterministic bus network and directly receives the time slot interrupt pulse issued by the deterministic bus network transmission cycle through the interrupt trigger register pin of the analog-to-digital converter, so that the sampling period is fixed at 0.01s and the clock phase deviation is less than 5μs. This constructs a hardware-aligned 4-point sliding sampling timing window for the communication cycle. With a working time of 0.01s, the operating current of the main drive motor of the screw press is continuously acquired at a sampling frequency of 100Hz. Temperature of the pressing chamber wall Current sampling time Obtained wall temperature The temperature was 110.5℃ at the previous sampling time. The wall temperature was 110.1℃, at the first two sampling times. The wall temperature was 109.7℃, at the first three sampling times. The wall temperature is 109.3℃. Historical temperature sequences are sequentially stored in a circular sliding buffer. Discrete weighted coefficients that decay over time are assigned to the wall temperature at the current sampling time, the previous sampling time, the second previous sampling time, and the third previous sampling time. The weighted discrete gradient is calculated, and the denoised wall temperature change rate is extracted. The speed is 40℃ / s, where, This is the temperature of the pressing chamber wall collected at the current sampling time. The temperature of the pressing chamber wall was collected at the previous sampling time. For the sampling period, in the actual temperature measurement arrangement, the temperature measurement point on the wall of the pressing chamber adopts a miniature fast-response thermocouple sensor. Its thermistor temperature measurement probe is directly embedded in the high shear friction micro-region that is attached to the inner wall of the pressing liner bar. The thermal response time constant of the temperature measurement probe is less than 5ms. When the local frictional resistance of the material in the pressing chamber increases sharply, the high-intensity shear friction power consumption between the material and the micro-region several micrometers deep on the surface of the inner liner metal is instantly converted into interfacial heat flow. The low thermal resistance of the shallow surface layer of the inner wall metal causes the transient heat flow to cause the surface temperature of the micro-region to rise within milliseconds. The wall temperature change rate obtained by weighted discrete gradient corresponds to the weighted discrete differential value of the transient friction heat generation rate of the surface of the shear micro-region.

[0042] The operating current of the main drive motor of the press screw is collected by the main drive control node at the current sampling time. The current is 48A, which is then substituted into the preset current conversion factor. The coefficient of temperature change is 0.25. It is 1.20, obtained through the formula. The pressing stress state parameters were calculated. It is 60, of which, For pressing stress state parameters, The current conversion factor, This refers to the operating current of the main drive motor of the screw press. The coefficient of temperature change. This is the temperature of the pressing chamber wall collected at the current sampling time. The temperature of the pressing chamber wall was collected at the previous sampling time. The sampling period is specified, and this value exceeds the preset stress disturbance threshold. A value of 20 indicates a non-linear accumulation trend of material within the pressing chamber. The main drive control node for pressing is then determined according to the formula... Calculate stress increment It is 40, of which, For stress increment, For pressing stress state parameters, The stress perturbation threshold is used; subsequently, a stress increment is generated. The differential state data frame with a deterministic timestamp is directly mapped into the deterministic real-time transmission time slot reserved in the deterministic bus network for masterless broadcasting, avoiding non-real-time data arbitration and queue buffering of other discrete nodes, and controlling the cross-node transmission delay of the bus to within 12ms.

[0043] The feed regulation control node and the cake discharge regulation control node synchronously receive differential state data frames from the deterministic bus network, and the feed regulation control node extracts the stress increment. Given a stress increment of 40, and combining this with the historical stress increment of 0 received in the previous frame, a discrete three-point backward difference operator algorithm with second-order truncation error accuracy is used based on the deterministic time within the data frame. This algorithm reads the current stress increment of 40, the previous historical stress increment of 0, and the historical stress increments of the two moments before that, and calculates the sum of 3 times the current stress increment minus 4 times the previous stress increment and 1 times the stress increment of the two moments before that. This sum is then divided by twice the fixed communication period time interval of 0.01s, yielding a first-order discrete rate of change of stress increment over time of 4000 per second. This result is then substituted into a preset gain scaling factor. The value is 0.15 and the differential adjustment coefficient. It is 0.02, according to the formula The amount of frequency reduction of the feed motor was calculated. It is 86Hz, where, The amount by which the operating frequency of the feed motor is reduced. This is the gain scaling factor. The differential adjustment coefficient is... For the received stress increment, To determine the rate of change of stress increment over time, the feed adjustment control node outputs a frequency reduction command, lowering the operating frequency of the feed motor by 86Hz from its current value. This reduces the speed of the feed auger and decreases the amount of material fed into the screw press feed section. Internally, it establishes a historical stress increment queue with a time-series depth of 3, sequentially recording the stress increment received in the current communication cycle, the historical stress increment of the previous communication cycle, and the historical stress increment of the two previous communication cycles. During differential calculation, it extracts a deterministic time interval of 0.01s between adjacent cycles based on the timestamps within the data frames. It subtracts four times the stress increment of the previous moment from three times the current stress increment, adds the stress increment of the two previous moments, and divides the resulting difference by twice the time interval to obtain the discrete rate of change. In the initial calculation phase of system initialization or abrupt change from zero, the historical stress increments of the previous moment and the two previous moments are both set to 0. This results in a stress increment rate of change of 4000 per second. When the rate of change of stress increment over time in subsequent communication cycles returns to 0 or a negative value, the feed adjustment control node gradually restores the operating frequency of the feed motor according to a preset step size.

[0044] The cake-dispensing adjustment control node extracts the stress increment from the differential state data frame. The pressure at the cake outlet and the position of the cake outlet mechanism are simultaneously acquired. The stress increment is multiplied by the preset cake outlet gap adjustment gain to calculate the cake outlet gap expansion amount of 0.5mm. The safe pressure threshold is pre-calibrated based on the allowable cake pressure range during continuous operation of the cake outlet mechanism. The maximum safe stroke limit is determined based on the allowable adjustment stroke of the cake outlet mechanism. When the cake outlet pressure exceeds the safe pressure threshold, the cake outlet adjustment control node controls the cake outlet gap adjustment module to expand the cake outlet gap by 0.5mm to release the physical pressure accumulated at the tail end of the pressing chamber. During the adjustment process, the position of the cake outlet mechanism is continuously acquired. When the position of the cake outlet mechanism reaches the maximum safe stroke limit, the expansion of the cake outlet gap is stopped, and a cake outlet limit alarm frame is sent to the deterministic bus network.

[0045] The steaming and frying control node continuously listens to differential state data frames on the deterministic bus network for a preset observation time window. To receive 10 consecutive stress increments within 2 seconds For differential state data frames greater than 0, the cumulative integral value within the time window is calculated. The value is 80, which is substituted into the preset steaming and frying integral compensation coefficient. It is 0.005, according to the formula The change in the opening degree of the steam valve for steaming and frying was calculated. It is 0.4%, of which, This represents the change in the opening degree of the steam valve for steaming and frying. This is the compensation coefficient for steaming and stir-frying. To observe the length of the time window, As the stress increment, the steaming and roasting control node outputs a valve adjustment signal, reducing the opening of the steaming and roasting valve by 0.4%. By adjusting the softening degree and moisture content of the rapeseed material inside the steaming and roasting pot, the accumulation of material in the pressing chamber is eliminated. When the pressing stress state parameter calculated by the pressing main drive control node... Falling back to the stress disturbance threshold When the value is below 20, the main drive control node of the pressing chamber stops broadcasting differential status data frames; after the feed adjustment control node, cake discharge adjustment control node, and steaming and roasting control node do not receive new differential status data frames within the preset timeout period, they respectively restore the operating frequency of the feed motor, the cake discharge gap, and the opening degree of the steaming and roasting valve to the preset closed-loop control baseline according to the preset linear slope, so that the pressure fluctuation range inside the pressing chamber is stabilized within the preset value. Within this period, screw jamming and main drive motor overcurrent protection failures no longer occur, and the residual oil content of the oil cake leaving the factory remains below 5.2%. In this cross-process collaborative control process, the system adopts a multi-time-scale hierarchical control mechanism. When a transient stress change occurs in the pressing chamber, the feeding adjustment control node and the cake discharge adjustment control node respectively execute the actions of reducing the feeding frequency and expanding the cake discharge gap within milliseconds to achieve physical pressure relief. The steaming and frying control node lowers the opening of the steam valve for long-cycle material plasticization state feedforward adjustment. After the heating amount is finely adjusted, the material embryo gradually changes the flow resistance and moisture content entering the pressing chamber during the subsequent conveying process, eliminating the continuous material accumulation trend from the material source end. The two types of adjustment form closed-loop control at different time scales.

[0046] Example 2: This example involves a 120-hour continuous operation test on a rapeseed oil pressing production line test bench. The test bench includes a steaming and roasting pretreatment unit, a feeding auger, a screw press unit, a cake outlet gap adjustment mechanism, and steaming and roasting control nodes, feeding adjustment control nodes, pressing main drive control nodes, and cake outlet adjustment control nodes connected via a deterministic bus network. The operating current of the screw press main drive motor is... The data was acquired using a Hall effect current sensor, with a measurement range of 0 to 100 A, a resolution of 0.01 A, and a sampling frequency of 100 Hz; the temperature of the pressing chamber wall was also measured. The system employs armored thermocouple sensors to collect data, with a measurement range of 0 to 200℃, a resolution of 0.05℃, and a sampling frequency of 100Hz. The cake discharge pressure is collected using a piezoelectric pressure sensor, with a measurement range of 0 to 20MPa and a resolution of 0.01MPa. The deterministic bus network uses a time-triggered deterministic bus scheduling protocol with a communication rate of 10Mbps. The reserved period for the deterministic real-time transmission time slot is 0.01s, and the maximum transmission delay limit is 12ms. A constant-temperature circulating water jacket is installed outside the pressing chamber of the screw press unit. A water temperature detection probe inside the jacket monitors the thermal load status of the outer wall of the pressing chamber. When the water temperature rises abnormally, the pressing main drive control node sends a speed limit command to the feed adjustment control node through the deterministic bus network to fine-tune the upper limit of the feed motor speed. Simultaneously, the material moisture content data measured by the steaming and roasting pretreatment unit is sent to the pressing main drive control node via the bus for verification with the spindle impedance torque signal calculated based on the motor current, used for cross-process feedforward correction.

[0047] During the experiment, Gaussian white noise and power frequency interference harmonics were injected into the sensor signal transmission channel to increase the operating current of the screw press main drive motor. The acquired signal is superimposed with ±1.8A random current noise, causing the temperature of the pressing chamber wall to rise. The acquired signal is superimposed with ±0.3℃ fluctuation thermal noise, and the sampling period is... Based on the response requirements and control node calculations for the stress mutation prediction in the pressing chamber, the load is set to 0.01s, and the number of points in the sliding sampling time window is... Indicates the number of consecutive sampling points participating in the calculation within the window; when When the value exceeds 6, the filter phase lag is greater than 0.05s. When the value is less than 3, it is difficult to eliminate single-point noise impact. Therefore, the residual minimization operation of weighted discrete gradient estimation is used to... Set to 4, stress disturbance threshold The envelope of the current fluctuation under no-load and steady-state load of the screw press was calibrated to 20. When the value was below this threshold, false broadcasts increased, and when the value was above this threshold, the warning time for sudden stress changes in the pressing chamber was shortened. During the entire test, the oil cake discharged from the cake outlet was sampled at fixed points every 2 hours and tested according to the standard residual oil determination method. The number of times the main drive motor stopped due to overcurrent was accumulated by the bottom log of the control system. When a differential state data frame was issued without actual mechanical impedance increase, it was recorded as a false broadcast. All test groups used the same batch of rapeseed raw material with a moisture content of 8.5%.

[0048] The test setup included one experimental group and six control groups. The experimental group employed a 4-point sliding sampling timing window and weighted discrete gradients, deterministic real-time transmission with masterless broadcasting within time slots, a deterministic second-order backward differential operator based on deterministic timestamps, and cross-node feedforward coordination. Control group one used a centralized master-slave control architecture, with a central master node aggregating sensor data and issuing control commands. Control group two used an independent single-loop PID control architecture, with each discrete node adjusting independently and no cross-node status data broadcasting. Control group three eliminated the 4-point sliding sampling timing window, directly using the temperature of adjacent sampling points for discrete difference. Control group four used a standard nondeterministic bus to transmit data frames and eliminated the deterministic second-order backward differential operator based on deterministic timestamps. Control group five used the current conversion coefficient... Set to 0.05, temperature change rate coefficient Set to 0.20; control group six will have a current conversion coefficient. Set to 0.80, temperature change rate coefficient Set it to 4.50.

[0049] In a typical material stacking condition, the main drive control node of the press continuously collects the operating current of the main drive motor of the screw press at a cycle of 0.01s. and the temperature of the pressing chamber wall Current sampling time The main drive motor of the screw press operates at a current of 48A, and the wall temperature is 110.5℃; at the previous sampling time... The first two sampling times and the first three sampling times The wall temperatures were 110.1℃, 109.7℃, and 109.3℃, respectively. The control group directly used discrete difference sampling from adjacent sampling points, obtaining a transient rate of change of 85℃ / s when noise peaks were superimposed. This resulted in the compression stress state parameter... A sudden increase in temperature triggered a false broadcast; the experimental group stored four consecutive wall temperature samples into a four-point sliding sampling time window, assigned discrete weighting coefficients that decayed over time, and calculated a weighted discrete gradient, obtaining a denoised wall temperature change rate of 40℃ / s. The experimental group then used the current conversion coefficient... Set to 0.25, temperature change rate coefficient Set to 1.20, the pressing stress state parameter is obtained according to the established calculation relationship between the operating current of the screw press main drive motor and the rate of change of wall temperature. The value is 60; this value is higher than the stress disturbance threshold. The stress increment is obtained from 20. The value is 40, and a differential state data frame containing the stress increment and a deterministic timestamp is generated and broadcast without a master through a deterministic bus network.

[0050] The feed regulation control node extracts the stress increment after receiving the differential state data frame. The value is 40, the historical stress increment of the previous frame is 0, and based on the deterministic timestamp within the data frame, the rate of change of stress increment over time is 4000 obtained using the deterministic time-based second-order backward difference operator; the gain scaling factor is... Set to 0.15, differential adjustment coefficient Setting the value to 0.02, a weighted sum of the stress increment and its rate of change over time is obtained to determine the reduction in the operating frequency of the feed motor. The frequency was 86Hz. Subsequently, the output frequency adjustment command reduced the speed of the feeding auger. When the control group four used ordinary nondeterministic bus transmission, the bus queue jitter was 25.4ms to 42.1ms. Due to the lack of deterministic timestamp alignment in the received data, the differential term corresponding to the second-order backward differential operator of the deterministic time base varied between 2100 and 5800, causing the feed motor operating frequency to decrease by ±18.2Hz, resulting in control amplitude oscillation.

[0051] The cake-dispensing adjustment control node synchronously receives differential state data frames and extracts stress increments. The pressure at the cake outlet and the position of the cake discharge mechanism are simultaneously acquired. The stress increment is multiplied by the preset cake discharge gap adjustment gain to calculate a cake discharge gap expansion of 0.5 mm. When the pressure at the cake outlet exceeds the preset safe pressure threshold, the cake discharge gap adjustment module expands the cake discharge gap by this expansion amount. The cake discharge gap adjustment actuator adopts a high-precision ball screw mechanism driven by a servo motor. The cake discharge mechanism includes a coaxial conical adjustment sleeve sleeved on the outside of the discharge cone of the screw press. The power output shaft of the servo motor is connected to the ball screw via a planetary reducer, which drives the conical adjustment sleeve to translate along the axis of the screw press to change the radial discharge gap between the conical adjustment sleeve and the discharge cone. A high-precision magnetostrictive linear displacement sensor is connected to the outer wall of the adjustment sleeve to provide real-time feedback on the position of the cake discharge mechanism. When the position reaches the preset maximum safe travel limit, the cake discharge adjustment control node stops expanding the cake discharge gap and sends a cake discharge limit alarm frame to the deterministic bus network.

[0052] The steaming and frying control node continuously listens to the differential state data frames in the deterministic bus network for the duration of the observation time window. To receive 10 consecutive stress increments within 2 seconds For data frames greater than 0, the cumulative integral value of 80 is calculated; this cumulative integral value is then compared with the steaming / frying integral compensation coefficient. Multiplying by 0.005 yields a 0.4% reduction in the steam valve opening for steaming and roasting, and an output valve adjustment signal is given to reduce the steam valve opening by 0.4% to adjust the softening degree of the rapeseed embryo. When the rate of change of stress increment over time returns to 0 or a negative value, the feed adjustment control node gradually restores the operating frequency of the feed motor according to the preset step size; when the material accumulation is eliminated and the pressing stress state parameter is... Falling back to the stress disturbance threshold When the following occurs, the pressing main drive control node stops broadcasting differential status data frames. After the feed adjustment control node, cake discharge adjustment control node, and steaming and frying control node do not receive new differential status data frames within the preset timeout period, they restore the feed motor operating frequency, cake discharge gap, and steaming and frying steam valve opening to the preset closed-loop control baseline according to the preset linear slope.

[0053] Continuous operation data were compared based on the differences in each control link. Under the complete scheme, the cross-node collaborative response time of the experimental group was 11.8ms, the false broadcast frequency was 0.2 times / h, the screw jamming rate was 0.0%, the residual oil content of the pressed cake was 5.12%, and the pressure fluctuation of the pressing chamber remained within ±1.2% of the preset value. After canceling the deterministic cross-node communication, the cross-node collaborative response time of control group one increased to 148.5ms, the false broadcast frequency was 1.1 times / h, the screw jamming rate was 14.2%, and the residual oil content of the pressed cake was 6.85%. Control group four, which used a common nondeterministic bus, had a cross-node collaborative response time of 38.6ms, a false broadcast frequency of 4.2 times / h, a screw jamming rate of 6.1%, and a residual oil content of the pressed cake of 5.92%.

[0054] Control group 2, which canceled cross-node status data broadcasting, had a cross-node collaborative response time of 165.2 ms, a false broadcast frequency of 0.0 times / h, a screw jamming rate of 22.5%, a residual oil content in the pressed cake of 7.42%, and a pressing chamber pressure fluctuation of ±12.4%. Control group 3, which canceled the 4-point sliding sampling timing window, had a cross-node collaborative response time of 12.1 ms, a false broadcast frequency of 28.4 times / h, a screw jamming rate of 8.3%, and a residual oil content in the pressed cake of 6.15%. When the coefficients deviated from the experimental group's set values, the control group... The cross-node collaborative response time of control group 5 was 12.0 ms, the false broadcast frequency was 0.1 times / h, the screw jamming rate was 18.6%, and the residual oil content of the pressed oil cake was 7.10%. The cross-node collaborative response time of control group 6 was 11.9 ms, the false broadcast frequency was 35.6 times / h, the screw jamming rate was 11.4%, and the residual oil content of the pressed oil cake was 7.82%. The above operating data correspond to the actual control results under deterministic communication, 4-point sliding sampling timing window, deterministic time-based second-order backward difference operator, cross-node feedforward collaboration, and parameter value changes, respectively.

[0055] Example 3: After a cold start or node reset in the rapeseed oil pressing process, the distributed control system operates on the physical production line, which includes steaming and roasting pretreatment, screw pressing, cake discharge and residue separation. The pressing main drive control node is connected to a deterministic bus network. During the initialization phase, the internal processor pre-fills the initial temperature sampling values ​​for four consecutive sampling cycles into a 4-point sliding sampling timing window; sampling cycle The operating current of the screw press main drive motor is collected at a sampling frequency of 100Hz with a time interval of 0.01s. Temperature of the pressing chamber wall Current sampling time The wall temperature was 110.5℃ at the previous sampling time. The wall temperature was 110.1℃, at the first two sampling times. The wall temperature was 109.7℃, at the first three sampling times. The wall temperature is 109.3℃. Four wall temperatures are sequentially stored in a ring-shaped sliding buffer and assigned discrete weighting coefficients of 0.40, 0.30, 0.20, and 0.10, respectively, decaying over time, according to the sampling time from nearest to farthest. The 0.10 value follows the system noise floor constant attenuation factor. A weighted discrete gradient is calculated based on the wall temperatures at the four sampling times. The denoised wall temperature change rate is 40℃ / s. The operating current of the screw press main drive motor at the current sampling time is 48A, and the current conversion coefficient... The coefficient of temperature change is 0.25. The value is 1.20, from which the pressing stress state parameter is obtained. The value is 60; this value exceeds the stress disturbance threshold. The stress increment was calculated from 20. The value is 40, and then a differential state data frame containing stress increments and deterministic timestamps is encapsulated and broadcast without a master via a deterministic bus network.

[0056] During the weighted discrete gradient calculation, the processor uses discrete weighting coefficients of 0.40, 0.30, 0.20, and 0.10 for 110.5℃, 110.1℃, 109.7℃, and 109.3℃ respectively, according to the order of the current sampling time, the previous sampling time, the second previous sampling time, and the third previous sampling time. Combined with a sampling period of 0.01s, the processor completes the weighted discrete gradient calculation within the 4-point sliding sampling time window, so that the wall temperature change rate after initialization is maintained at 40℃ / s.

[0057] The feed regulation control node and the cake discharge regulation control node receive differential state data frames from the deterministic bus network. The feed regulation control node extracts the stress increment 40, reads the historical stress increment 0 from the previous frame, obtains the communication time interval based on the deterministic timestamp in the data frame, and calculates the rate of change of stress increment over time 4000 using a deterministic time-based second-order backward differential operator; gain scaling factor. The differential adjustment coefficient is 0.15. The value is 0.02, from which the reduction in the operating frequency of the feed motor is obtained. The calculated frequency is 86Hz, which exceeds the maximum single-step frequency reduction limit of 60Hz. Therefore, the frequency reduction of the feeding motor is cut off and locked to 60Hz, and then a frequency reduction command is output to reduce the speed of the feeding auger to limit the mechanical impact caused by a single speed reduction. The feeding motor is a wide-frequency variable frequency motor with a rated operating frequency of 100Hz. The reference operating frequency during normal steady-state operation is 80Hz to 90Hz. When the calculated frequency reduction requirement reaches 86Hz, it is cut off to 60Hz by the maximum single-step frequency reduction limit, so that the motor operating frequency is kept in a low-frequency feeding state of 20Hz to 30Hz.

[0058] The cake discharge adjustment control node synchronously extracts the stress increment 40 from the differential state data frame, and obtains the cake discharge outlet pressure and the position of the cake discharge mechanism. It multiplies the stress increment by the preset cake discharge gap adjustment gain to calculate the cake discharge gap expansion amount of 0.5mm. When the cake discharge outlet pressure exceeds the preset safe pressure threshold, the control module expands the cake discharge gap by 0.5mm to release the pressure accumulated at the end of the pressing chamber. During the process of expanding the cake discharge gap, the position of the cake discharge mechanism is continuously monitored. When the position reaches the preset maximum safe stroke limit, the expansion of the cake discharge gap is stopped, and a cake discharge limit alarm frame is sent to the deterministic bus network.

[0059] The steaming and frying control node continuously listens to the differential state data frames in the deterministic bus network for the duration of the observation time window. After receiving 10 consecutive data frames with stress increments greater than 0 within 2 seconds, the cumulative integral value of 80 was calculated; steaming and frying integral compensation coefficient. The value is 0.005, from which the change in the opening degree of the steam valve for steaming and frying is obtained. The value is 0.4%, and a valve adjustment signal is output to reduce the opening of the steam valve for steaming and frying by 0.4% to adjust the softening degree and moisture content of the rapeseed material.

[0060] When the rate of change of stress increment over time returns to 0 or a negative value, the feed adjustment control node gradually restores the operating frequency of the feed motor according to the preset step size, and when the pressing stress state parameter Continue to fall back to the stress disturbance threshold When the value is below 20, the main drive control node of the pressing machine stops broadcasting differential status data frames; after the feed adjustment control node, cake discharge adjustment control node and steaming and frying control node do not receive a new differential status data frame within the preset timeout period of 1 second, they respectively restore the operating frequency of the feed motor according to the preset linear slope of 2Hz / s, restore the cake discharge gap according to the preset linear slope of 0.05mm / s, and restore the opening of the steaming and frying valve according to the preset linear slope of 0.05% / s, until they are restored to the preset closed-loop control baseline.

[0061] Example 4: This example combines Figures 1 to 2 This paper describes a decentralized control system for rapeseed oil pressing, such as... Figure 1As shown, the distributed control system includes a steaming and roasting control node, a pressing main drive control node, a deterministic bus network, a feed regulation control node, and a cake discharge regulation control node. The steaming and roasting control node is connected to both a steaming and roasting temperature acquisition module and a discharge humidity acquisition module. It collects the rapeseed steaming and roasting temperature and discharge humidity and outputs steam valve control signals. It also connects to the deterministic bus network to listen for data frames. The pressing main drive control node is connected to a main drive current acquisition module, a wall temperature acquisition module, and a stress parameter calculation module. It collects the main drive motor operating current and the pressing chamber wall temperature, constructs a 4-point sliding sampling timing window with a communication cycle locked to the deterministic bus network, calculates the denoised wall temperature change rate using the wall temperature sequence of four consecutive sampling times, and calculates the pressing stress state parameter. Furthermore, when the pressing stress state parameter exceeds a preset stress... When the disturbance threshold is reached, a differential state data frame containing stress increment and timestamp is generated and broadcast without a master to the feed adjustment control node and the cake discharge adjustment control node via a deterministic bus network. The deterministic bus network maintains a communication connection with each control node for transmitting and forwarding data frames. The feed adjustment control node is connected to the feed drive module and the speed monitoring module to collect the operating frequency of the feed motor and the speed of the feed mechanism and output the feed drive command. At the same time, it receives data frames from the deterministic bus network. The timestamp is calculated using a deterministic time-based second-order backward difference operator to calculate the rate of change of stress increment over time and to lower the operating frequency of the feed motor. The cake discharge adjustment control node is connected to the cake discharge gap adjustment module to receive data frames from the deterministic bus network and to control the cake discharge gap adjustment module to increase the cake discharge gap according to the stress increment.

[0062] like Figure 2 As shown, in the system's collaborative control logic, the pressing main drive control node constructs a 4-point sliding sampling timing window to calculate the rate of change of the wall temperature after noise reduction, without a main broadcast differential state data frame. The feeding adjustment control node uses a deterministic time-based second-order backward differential operator to calculate the rate of change of stress increment over time, thus mitigating the mechanical impact caused by the deceleration of the feeding mechanism. The deterministic bus network avoids arbitration preemption and queue buffering jitter by opening a periodic cyclic time slot sequence, controlling the transmission delay within a preset upper limit. The cake discharge adjustment control node extracts the cake discharge pressure and mechanism position, controls the cake discharge gap adjustment module to expand the gap, and sends a cake discharge limit alarm frame when the limit is reached. The steaming and frying control node continuously receives data frames indicating that the stress increment is greater than zero, calculates the cumulative integral value of the stress increment within the time window, and reduces the opening of the steaming and frying steam valve. Thus, multiple nodes and the deterministic bus network work together to improve the control stability and continuity of the continuous pressing process.

[0063] Example 5: To address the baseline drift in pressing caused by fluctuations in the moisture content of different batches of rapeseed raw materials, the distributed control system reads the no-load operating current of the screw press and the ambient temperature data of the pressing chamber through the main drive control node before entering the continuous pressing process. Under the condition that the screw press runs no-load for 300 seconds without material conveying, the operating current of the main drive motor of the screw press is collected at a sampling frequency of 100Hz. The baseline of the no-load current was obtained; then, thermal excitation was applied to the module inside the pressing chamber, and the thermal response decay curve of the wall temperature was obtained. The thermal conduction decay constant was derived by least squares fitting, and the current conversion coefficient was... The coefficient of temperature change is calibrated to 0.25. The stress disturbance threshold was calibrated to 1.20, and based on the measured initial moisture content of the rapeseed raw material, the stress disturbance threshold was set accordingly. The stress disturbance threshold is set to 20. When the moisture content of the raw material is higher than 12.0%, the stress disturbance threshold will be adjusted. A slight upward adjustment of 10.0% compensates for the increase in steady-state current caused by the flow resistance of wet materials, thus adjusting the pressing stress state parameters. Maintain stress disturbance threshold when there is no material buildup. the following.

[0064] The pressing stress state parameter is used to characterize the comprehensive mechanical impedance and thermodynamic load inside the pressing chamber, and is a dimensionless normalized index. The main drive motor operating current reflects the electromagnetic torque required for the main shaft to overcome the axial propulsion resistance of the material, and the wall temperature change rate reflects the transient heat energy release rate caused by high shear friction. In the coefficient calibration process, the no-load baseline current of the motor is first measured under no-load conditions, and the reciprocal of the rated operating current range is converted into a current conversion coefficient of 0.25. Then, a step load with a known gradient is applied to the no-load pressing chamber, and the proportional response relationship between the shear temperature rise slope and the main shaft impedance torque is measured. The temperature change rate coefficient of 1.20 is derived and used for the dimensionless normalization fusion of the electromagnetic torque parameter and the frictional heat flow parameter.

[0065] When each node starts up, the deterministic bus network initiates deterministic frame timing synchronization with the pressing master drive control node as the clock master source, and broadcasts time step pulses with high-precision hardware timestamps to the steaming and roasting control node, the feeding adjustment control node, and the cake discharge adjustment control node. After receiving the pulses, each slave node calculates the bus physical propagation delay and the local crystal oscillator drift deviation, locking the discrete clock phase error of each node within 5.0μs within a 0.01s communication cycle. When offline data backup or non-real-time parameter download is enabled during the pressing process, the deterministic bus network restricts the corresponding non-real-time data frames to be sent within the non-deterministic non-real-time transmission time slot, without occupying the slot containing stress increments. The deterministic real-time transmission time slot reserved for the differential state data frame keeps the masterless broadcast delay of the differential state data frame within the 12ms limit.

[0066] The scheduling cycle of the deterministic bus network is set to 20ms, and symmetrically divided into two 10ms discrete communication sub-cycles. The first 2ms of each 10ms sub-cycle is a deterministic real-time transmission time slot dedicated to broadcasting differential state data frames, and the following 8ms is a non-deterministic non-real-time transmission time slot. The 12ms transmission delay limit includes the data acquisition and calculation time, frame encapsulation time, and the maximum physical end-to-end delay required for data to cross into the next sub-cycle's deterministic real-time transmission time slot to complete broadcast reception. Each data frame carries a nanosecond-level hardware timestamp, and each slave node performs pipelined time base alignment in the receive buffer based on this timestamp, so that differential operations are always performed based on a deterministic time base.

[0067] Example 6: When multiple parallel-configured screw press units are connected to the rapeseed oil pressing production line, each screw press unit is connected to a pressing main drive control node and a feed adjustment control node. Each pressing main drive control node shares its own main drive motor operating current derivative in real time through a deterministic bus network in a peer-to-peer communication manner. One of the screw press units experienced a sudden increase in the main drive motor's operating current from 48A to 56A due to abnormal feed particle size. The current reaches 0.8A / s. When this value remains higher than the average current derivative of 0.2A / s for the other parallel screw press units for 0.05s, the corresponding main drive control node generates a current shunting adjustment data frame and broadcasts it without a master through the deterministic bus network. The current shunting adjustment module receives the current shunting adjustment data frame, extracts the current derivative deviation of 0.6A / s, substitutes it into the current shunting gain coefficient to calculate a 15% reduction ratio in the current shunting flow. The opening of the feed valve of the overloaded screw press unit is reduced by 15%. At the same time, according to the current shunting sharing rule, the opening of the feed valve of the other screw press units whose main drive motor operating current derivative is within the normal range is synchronously increased by 7.5%. The normal range is determined according to the range of the operating current derivative of the main drive motor when each screw press unit is not overloaded.

[0068] The feed distribution and regulation module includes a multi-port distribution manifold located below the main raw material feed pipe, and electrically operated regulating butterfly valves installed on each manifold. Each electrically operated regulating butterfly valve receives flow distribution and regulation commands via a servo controller and changes the valve plate opening. The oil collection tank filter screen at the oil outlet self-cleaning node uses a double-layer stainless steel woven filter screen, with piezoresistive pressure sensors installed on the upper and lower sides of the filter screen to detect the pressure difference across the oil collection tank filter screen. The sludge removal and cleaning module includes a low-speed rotating spiral scraper that is in close contact with the surface of the filter screen, and a high-pressure electromagnetic pulse nozzle connected to the high-temperature oil cleaning and rinsing circuit. The high-pressure jet, in conjunction with the rotation of the scraper, removes the oil sludge trapped on the filter screen.

[0069] At the oil outlet plate location, the self-cleaning node collects the pressure difference across the oil collection tank filter screen. With oil outlet pipeline flow rate And calculate the rate of change of pressure difference. With the rate of decrease in oil outlet flow When the static pressure of oil residue builds up on the filter screen, it causes a pressure difference between the two sides of the filter screen in the oil collection tank. As the pressure rises to 0.15 MPa, the pressure differential increases. The pressure reaches 0.05 MPa / s, while the flow rate in the oil outlet pipeline decreases. When the flow rate reaches -0.2L / s², the oil outlet self-cleaning node outputs a drive signal to increase the operating frequency of the spiral slag scraper from 0.1Hz to 0.5Hz, and starts the slag discharge flushing circuit to perform a 10-second cycle flushing to remove the oil sludge accumulated on the filter screen.

[0070] Temperature of the pressing chamber wall is collected at the pressing main drive control node. During the process, if the wall temperature suddenly drops to 0℃ due to a disconnection in the acquisition channel, or if the acquired value exceeds the physical measurement limit of 180℃, the wall temperature acquisition channel is deemed faulty, and the current wall temperature is discarded. The system samples the wall temperature change rate at the previous valid moment and locks it in; simultaneously, it broadcasts sensor alarm data frames to the deterministic bus network to prevent abnormal temperature sampling values ​​from affecting the pressing stress state parameters. False stress disturbance threshold To maintain stable operation of the system during physical hardware failure maintenance.

[0071] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0072] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A decentralized control system for rapeseed oil pressing, characterized in that, Include: The steaming and frying control node includes a steaming and frying temperature acquisition module and an output humidity acquisition module, which acquires the rapeseed steaming and frying temperature and output humidity, and outputs steam valve control signals for steaming and frying. The feeding adjustment and control node includes a feeding drive module and a speed monitoring module, which collects the operating frequency of the feeding motor and the speed of the feeding mechanism, and outputs feeding drive commands. The main drive control node for pressing includes a main drive current acquisition module, a wall temperature acquisition module, and a stress parameter calculation module. It acquires the operating current of the main drive motor and the wall temperature of the pressing chamber, constructs a 4-point sliding sampling timing window with a communication cycle locked with the deterministic bus network, calculates the denoised wall temperature change rate using the wall temperature sequence of 4 consecutive sampling times, and calculates the pressing stress state parameters. When the pressing stress state parameters exceed the preset stress disturbance threshold, it generates a differential state data frame containing stress increment and timestamp, and broadcasts it to the feed adjustment control node and the cake discharge adjustment control node without a master through the deterministic bus network. The feed adjustment control node receives differential state data frames, calculates the rate of change of stress increment over time using a deterministic time-based second-order backward differential operator based on the timestamp, and lowers the operating frequency of the feed motor. The cake discharge adjustment control node includes a cake discharge gap adjustment module, which receives differential state data frames and controls the cake discharge gap adjustment module to expand the cake discharge gap according to the stress increment.

2. The decentralized control system for rapeseed oil pressing according to claim 1, characterized in that, The deterministic bus network adopts a time-triggered deterministic bus scheduling protocol. Within the communication cycle, the deterministic bus network opens a periodic cyclic sequence of time slots and divides the time slot sequence into deterministic real-time transmission time slots and non-deterministic non-real-time transmission time slots. The master drive control node directly maps the differential state data frame into the reserved deterministic real-time transmission time slot for masterless broadcasting, avoiding arbitration and queue buffering jitter when transmitting non-real-time data generated by other discrete nodes on the bus, and controlling the transmission delay of the differential state data frame on the bus within the preset upper limit of delay.

3. The decentralized control system for rapeseed oil pressing according to claim 1, characterized in that, The feed adjustment control node calculates the amount of frequency reduction of the feed motor by weighted summing of the extracted stress increment and its rate of change over time based on the preset gain ratio coefficient and differential adjustment coefficient. When the frequency reduction of the feeding motor exceeds the preset maximum frequency reduction limit, the feeding adjustment control node will cut off and lock the frequency reduction of the feeding motor to the maximum frequency reduction limit and output a frequency reduction command to the feeding drive module to suppress the mechanical impact caused by the deceleration of the feeding mechanism. After the rate of change of stress increment over time returns to zero or a negative value, the feed adjustment control node gradually restores the operating frequency of the feed motor according to the preset step size.

4. A decentralized control system for rapeseed oil pressing as described in claim 1, characterized in that, The steaming and frying control node connects to the deterministic bus network and listens to differential state data frames; when it continuously receives differential state data frames indicating that the stress increment is greater than zero within the preset observation time window, the steaming and frying control node calculates the cumulative integral value of the stress increment within the observation time window; The steaming and frying control node calculates the amount of reduction in the steam valve opening based on the product of the accumulated integral value and the preset steaming and frying integral compensation coefficient, and then reduces the opening of the steam valve.

5. A decentralized control system for rapeseed oil pressing as described in claim 1, characterized in that, When the cake discharge adjustment control node receives the differential state data frame, it extracts the cake discharge port pressure and the position of the cake discharge mechanism; the cake discharge adjustment control node multiplies the stress increment with the preset cake discharge gap adjustment gain to calculate the cake discharge gap expansion amount; When the pressure at the cake outlet exceeds the preset safety pressure threshold, the cake outlet adjustment control node controls the cake outlet gap adjustment module to expand the cake outlet gap by the amount of cake outlet gap expansion; when the position of the cake outlet mechanism reaches the preset maximum safety travel limit, the cake outlet adjustment control node stops expanding the cake outlet gap and sends a cake outlet limit alarm frame to the deterministic bus network.

6. A decentralized control system for rapeseed oil pressing according to claim 1, characterized in that, When constructing a 4-point sliding sampling time window, the main drive control node of the pressing chamber stores the wall temperature of the pressing chamber collected in 4 consecutive sampling cycles in a ring-shaped sliding buffer. The main drive control node of the pressing chamber assigns discrete weighting coefficients that decay with time to the wall temperature at the current sampling time, the wall temperature at the previous sampling time, the wall temperature at the second sampling time, and the wall temperature at the third sampling time, calculates the weighted discrete gradient, and extracts the wall temperature change rate.

7. A decentralized control system for rapeseed oil pressing according to claim 1, characterized in that, When the pressing process includes multiple screw press units configured in parallel, each screw press unit is connected to a pressing main drive control node and a feed adjustment control node respectively; each pressing main drive control node shares its own main drive motor operating current derivative in real time through a deterministic bus network in a peer-to-peer communication manner. When the operating current derivative of a certain screw press unit is continuously higher than the average current derivative of the other screw press units, the corresponding main drive control node generates a current shunting adjustment data frame and outputs a current shunting adjustment command to the material distribution adjustment module, reducing the current shunting flow to that screw press unit and increasing the current shunting flow to the other screw press units whose operating current derivatives are within the normal range.

8. A decentralized control system for rapeseed oil pressing as described in claim 1, characterized in that, The distributed control system also includes an oil outlet self-cleaning node; the oil outlet self-cleaning node collects the pressure difference on both sides of the oil collection tank filter screen of the oil outlet pan and the flow rate of the oil outlet pipeline, and calculates the transient rate of change of the flow rate and pressure difference of the oil outlet pipeline; when the rate of increase of the pressure difference and the rate of decrease of the flow rate both exceed their respective preset thresholds, the oil outlet self-cleaning node controls the slag discharge cleaning module to adjust the action frequency and flushing cycle.

9. A decentralized control system for rapeseed oil pressing according to claim 1, characterized in that, When the pressing stress state parameter calculated by the pressing master drive control node falls below the stress disturbance threshold, the pressing master drive control node stops broadcasting differential state data frames to the deterministic bus network. If the feeding adjustment control node, the cake discharge adjustment control node, and the steaming and frying control node do not receive the differential state data frame within the preset timeout period, the feeding motor operating frequency, the cake discharge gap, and the steaming and frying valve opening shall be restored to the preset closed-loop control baseline according to the preset linear slope.

Images