Self-adaptive high-precision discharging control method for materials

Abstract

The invention relates to a material self-adaptive high-precision discharging control method. The method comprises a learning training process and a normal discharging process, wherein the learning training process comprises the steps of collecting data; manual operation; learning training; the normal discharging process comprises the steps that a gate is opened for discharging; material lowering monitoring is carried out; controlling the lowering flow; and discharging is finished. The gate is controlled through PID, meanwhile, a dead zone of the gate is recognized and determined in advance in a learning machine mode, and therefore the problem that the loading capacity is difficult to control when the gate of a multi-bucket rotary loading system is closed is solved, self-adaptive control over the gate is achieved according to material property changes in the discharging process, and the loading capacity of the multi-bucket rotary loading system is improved. And the discharging weight control precision is ensured. The control process can also be applied to the warehouse feeding process of a conventional loading station, and self-adaption of material properties is achieved.

Description

Technical Field

[0001] This invention relates to a method for material feeding under adaptive high-precision control, a mechanical automated transportation method, and an automated loading method for bulk material transportation. Background Technology

[0002] Multi-bucket rotary loading systems are a type of highly automated and efficient bulk material loading equipment that has emerged in recent years. This type of loading equipment has the ability of traditional loading stations to quickly and integratedly distribute, weigh, and unload materials using gravity. It also has the advantages of a lower overall height, reducing the height of material lifting and steel structure frame, resulting in significant overall energy savings. In addition, it occupies less space and is suitable for use in places with limited space, such as docks.

[0003] Existing multi-bucket rotary loading systems sometimes employ a station-based method for material feeding control. During feeding, the feeding area is located at the center of the bin. Once the weight requirement of the material in the bin is met, the system rotates to the next station to feed the next bin. Although the first station meets the feeding value, feeding does not stop during rotation. Moving from the center of one bin to the center of another, a span of 120 degrees, requires motor acceleration and gradual deceleration. During loading, the conveyor belt is constantly feeding. Therefore, material continues to enter the bin during bin changes. Consequently, the actual material value in the bin exceeds the target material value. When the conveying speed is high, the weight in the previous bin may far exceed the target weight after a bin change. Therefore, it is necessary to control the material weight during loading by controlling the discharge. In this case, precise feeding of material in the multi-bucket bins is not required; the amount of material is mainly controlled by the discharge. In the process of precise material feeding control, changes in material flowability have a significant impact on the accuracy of feeding control. How to achieve precise control of the fed material is a problem that needs to be solved. Summary of the Invention

[0004] To overcome the problems of existing technologies, this invention proposes a method for high-precision adaptive material feeding control. The method utilizes PID control to regulate the gate, and employs a learning machine to identify the gate dead zone and confirm the lead time, thereby solving the problem of difficulty in controlling the loading volume when the gate closes in a multi-bucket rotary loading system.

[0005] The objective of this invention is achieved as follows: a method for material feeding under adaptive high-precision control, wherein the loading station using the method is a multi-bucket rotary loading system, and the method includes a learning and training process and a normal feeding process.

[0006] The learning and training process includes the following steps:

[0007] Step 101, Data Collection: Collect information on the discharge of various bulk materials in the past, including the flowability of the materials, the size of the dead zone, the amount of lead time, and the parameters of gate closure;

[0008] Step 102, Manual Operation: Manually operate the material feeding process using materials with various characteristics, and collect various gate parameters during the operation, including material flowability, dead zone size, lead time, and gate closing parameters; among which, flowability is determined based on the monitoring of material movement and material accumulation by sensors during the actual operation.

[0009] Step 103, Learning and Training: Input the collected data and manual operation data into the database, and train the gate to close using a learning machine, so as to achieve adaptive gate closing to dead zone and lead time; the database will be continuously improved in the subsequent normal material feeding process;

[0010] The normal material feeding process includes the following steps:

[0011] Step 201, Open the gate to release material: After receiving the release command, first record the weight value of the material at the highest position in the hopper. After the unloading begins, the gate is initially fully open.

[0012] Step 202, Material release weight monitoring: Weight monitoring is performed in real time through the weighing sensors in the hopper;

[0013] Step 203, Material discharge flow control: The flow rate is monitored in real time by the gate opening, the amount of material flowing out, and the changes in the accumulation of material in the carriage;

[0014] PID monitors the gate opening process using the following formula:

[0015]

[0016] Among them, the size of the dead zone and the size of the lead are given based on the learning results of the learning machine;

[0017] Step 204, End of material feeding: The controller obtains the current dead zone and lead time of the loading based on the learning and training. The gate stops at the lead time position, and the material continues to be fed down slowly until the target value is reached. The gate then closes completely, and the feeding ends.

[0018] The advantages and beneficial effects of this invention are as follows: This invention utilizes PID control to regulate the gate, and simultaneously employs a learning machine to identify the gate dead zone and confirm the lead time, thereby solving the problem of difficulty in controlling the loading amount when the gate closes in a multi-bucket rotary loading system. It achieves adaptive control of the gate based on changes in material properties during the unloading process, ensuring the accuracy of unloading weight control. This control process can also be applied to the binning process in conventional loading stations to achieve adaptive material property control. Attached Figure Description

[0019] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0020] Figure 1 This is a schematic elevation view of the loading station to which the material feeding method described in this embodiment of the invention is applied;

[0021] Figure 2 This is a schematic diagram of the loading station structure to which the material feeding method described in this embodiment of the invention is applied;

[0022] Figure 3 This is a block diagram of the control principle of the loading station used in the material feeding method described in the embodiments of the present invention;

[0023] Figure 4 This is a schematic diagram of the PID control algorithm for the feeding method described in this embodiment of the invention;

[0024] Figure 5 This is a schematic diagram of the gate dead zone and lead time of the loading station used in the material feeding method described in this embodiment of the invention. Figure 1 Enlarged view of point F in the middle;

[0025] Figure 6 This is a schematic diagram illustrating the process of tracking the weight and flow rate of materials in the silo using the material feeding method described in this embodiment of the invention.

[0026] Figure 7 This is a flowchart of the material feeding method described in an embodiment of the present invention. Detailed Implementation

[0027] Example:

[0028] This embodiment describes a method for adaptive high-precision control of material feeding. The loading station used in this method is a multi-bucket rotary loading system. Figure 1 , 2 As shown.

[0029] The multi-bucket rotary loading system is a loading station where multiple hoppers are arranged at the same horizontal level. These hoppers simultaneously function as buffers, weighers, and unloaders. Multi-bucket rotation means dividing a single cylindrical silo into three 120-degree sector-shaped hoppers 1. The three hoppers (A, B, and C) are relatively independent, each with its own weighing device 2 and its own unloading gate 3. The three hoppers share a common inlet 4 and a chute 5. During loading, the three hoppers rotate horizontally (…). Figure 2 (In the direction of arrow E), the belt conveyor continuously feeds material into one of the rotating hoppers, while the other two hoppers either unload or await unloading. The three hoppers act as buffers for each other, and the unloading volume is monitored by a weighing device and a flow monitoring device. During the rotation of the hoppers, the three hoppers sequentially complete the actions of feeding, weighing, and unloading. While the hoppers are rotating, the carriage either stops or moves forward. Figure 1 , 2 (In the direction of arrow D), the material is piled up in the car body. The factors involved in controlling the unloading include: adjusting the rotation of the three hoppers, adjusting the conveyor belt's material feeding, and adjusting the gate opening. Therefore, the unloading control subsystem includes: a rotary variable frequency motor driving the overall rotation of the three hoppers, unloading gates and opening sensors for the three hoppers, weighing devices for the three hoppers, a flow monitoring device, a variable frequency motor for the belt conveyor, car body position and speed measuring gratings, and a controller with PID control, a database, and an intelligent learning machine, such as... Figure 3 As shown.

[0030] A crucial function of any automated loading station is the accurate weighing of materials, thereby precisely controlling the weight of materials loaded (loading quantity) to meet the needs of commercial operations. Since the loading quantity is essentially the delivery quantity, it is a fundamental and essential element of any commercial contract and must be guaranteed.

[0031] There are two ways to control the weight of materials being loaded in a multi-bucket rotary loading system:

[0032] One method is to precisely control the amount of material entering the hopper, minimizing the error between the actual amount of material entering the hopper and the expected amount of material (i.e., the planned amount of material to be loaded, or the target amount of material). When unloading, simply open the gate to unload all the material in the hopper to achieve a precise loading amount. This method can be called the material loading control method.

[0033] Another method is to precisely control the amount of material discharged, which involves pre-loading the hopper with more material than expected, and then controlling the gate during discharge to minimize the error between the actual loading amount and the expected material amount. This method can be called the discharge control method. This embodiment adopts the discharge control method.

[0034] In this embodiment, the belt conveyor continuously feeds material into the hopper. The amount of material in the hopper changes dynamically in real time, and there are dead zone issues when the gate closes, as well as issues related to improving the accuracy of material discharge. This embodiment addresses these two issues separately. First, the traditional PID control of the unloading gate is modified. Although PID control can achieve considerable control accuracy, it is insufficient in some demanding loading scenarios. Therefore, an intelligent learning approach is needed to further improve control accuracy and effectively control the loading volume. The intelligent learning machine is an electronic system such as an artificial neural network or Bayesian network that has the ability to learn intelligently from past experience to solve problems.

[0035] The gate position control algorithm based on PID algorithm: The gate equipped in the silo is controlled by a hydraulic proportional valve.

[0036] In practical application, the target angle is used as the target engineering quantity for PID control. The actual angle of the multi-stage oscillator is used as the input, and the target value is subtracted from the actual angle value.

[0037] The specific PID formula is as follows:

[0038]

[0039] This allows for precise control of the gate position via the current flowing through the proportional valve, such as... Figure 4 As shown.

[0040] Using the PID algorithm described above, the gate position can be precisely controlled, with an accuracy within ±1cm. Under ideal conditions, the relationship between the gate position and the material discharge is as follows:

[0041]

[0042] The gate position changes as the material is fed in. Initially, the weight value at the highest point in the weighing bin is recorded. After unloading begins, the gate is initially fully open. As material is fed in, the weight change in the bin is tracked. By subtracting the percentage of the remaining material in the bin from the target feeding amount and multiplying this by the fully closed position, a linear relationship can be established between the gate's closing degree and the feeding value. Finally, when the feeding reaches the target value, the gate closes simultaneously, achieving high-precision feeding control.

[0043] The PID control algorithm described above can effectively adapt the gate closing speed according to the material's flowability. When the material has good flowability, the gate closes quickly; when the material has poor flowability, the gate closes slowly.

[0044] However, the above represents an ideal scenario. This is because when the gate 301 closes to a certain extent, the material cannot fall. The inability to fall prevents the gate from closing, thus halting further operation and creating a dead zone. Figure 5 As shown.

[0045] To avoid dead zones, the PID control algorithm needs to be modified. The modification formula is as follows:

[0046]

[0047] As shown in the above formula, the controlled position does not ultimately need to be completely closed; the action stops after reaching a certain position to ensure the material falls. The falling amount minus a lead time (see...) Figure 5 The gate is fully closed only when the target value is reached. The purpose of the lead time is to ensure that material will continue to fall at the moment the gate is fully closed, so that the amount of material released after the gate is closed is closer to the target value.

[0048] Adaptive verification of dead zone and lead: When dead zone and lead are added to the control, it is necessary to verify the values ​​of dead zone and lead.

[0049] The size of the dead zone is related to the material characteristics. Materials that tend to accumulate will have a larger dead zone; materials with poor flowability will also have a relatively large dead zone. Furthermore, the size of the dead zone directly determines the final cutting accuracy. For example, materials with good flowability and small particle size should have a smaller dead zone.

[0050] Lead time also affects the accuracy of material feeding control. Materials with good flowability require a larger lead time, while materials with poor flowability require a smaller lead time. Conventionally, this lead time varies depending on the material. This means that when material properties change, a human must input the lead time as a parameter into the system based on experience to adapt to the changes in different materials. This is a major reason why conventional methods cannot achieve adaptive control.

[0051] During the gate control process, after the gate is properly adjusted, the closing of the dead zone is instantaneous and the time is basically fixed.

[0052] To achieve adaptive dead zone and lead time control based on material properties, this embodiment employs an intelligent learning approach. First, during the manual operation learning process, the lead time is set to a constant (referred to as the lead time constant). When the lead time is set to a constant, the cut-off value corresponding to the gate closing at the lead time is fixed within the allowable error range (the cut-off value refers to the timing of the valve closing the material quantity during material release; its error range is the difference between the material pile quantity at the time of cut-off and the target material pile quantity). Therefore, only the size of the dead zone needs to be adjusted according to the material's flowability to achieve adaptive control over material properties. However, the relationship between the lead time constant and the error of the cut-off value is non-linear, making it difficult to perfectly correspond using a fixed constant. Therefore, the lead time constant needs to be set within a certain range to minimize the cut-off value error. The intelligent learning method digitizes the experience gained during manual loading and previous loading experiences, training the controller to automatically select the appropriate lead time constant based on the material characteristics and on-site loading conditions, minimizing the cut-off value error and achieving precise material release and loading.

[0053] During the unloading process, in addition to tracking the weight of the material in the silo, it is also necessary to track the flow rate of the material. Both tracking methods can be achieved in existing instruments, and the tracking process is as follows: Figure 6 As shown.

[0054] When the material is discharged, the weight and flow rate of the material in the silo are tracked.

[0055] The gate's stroke is adaptively controlled based on changes in the weight of the material in the silo. The gate gradually closes as material is released, and the flow rate gradually decreases as the gate closes.

[0056] A minimum flow rate is set. When the flow rate decreases to this minimum value, the system is notified to stop gate travel control based on material weight changes. The gate will stop at a position, leaving a dead zone length.

[0057] The material continues to fall, and when it reaches the allowable advance value, the gate is closed instantly to complete the material discharge.

[0058] When the properties of a material change, its flowability also changes. When adaptive control is used to address these changes, it primarily affects the size of the final dead zone. However, under final dead zone control, the material flow rate and lead time remain consistent, thus ensuring uniformity in material feeding and cutting regardless of material variations, achieving the same flow rate and lead time.

[0059] The material feeding method under adaptive high-precision control described in this embodiment includes a learning and training process and a normal feeding process, as follows: Figure 7 As shown:

[0060] The learning and training process includes the following steps:

[0061] Step 101, Data Collection: Collect information on the discharge of various bulk materials in the past, including the material flowability and dead zone size, lead time, and gate closing parameters.

[0062] In highly automated loading stations, various loading parameters are typically recorded during the loading process of various bulk materials. Extracting useful data from these records is an essential task.

[0063] Step 102, Manual Operation: Manually operate the material feeding process using materials with various characteristics, and collect various gate parameters during the operation, including material flowability, dead zone size, lead time, and gate closing parameters; among which, flowability is determined based on the monitoring of material movement and material accumulation by sensors during actual operation, rather than simply extracted from previous data.

[0064] Step 103, Learning and Training: Input the collected data and manual operation data into the database, and train the gate to close using a learning machine, so as to achieve adaptive gate closing to dead zone and lead time; the database will be continuously improved in the subsequent normal material feeding process.

[0065] The normal material feeding process includes the following steps:

[0066] Step 201, Open the gate to release material: After receiving the release command, first record the weight value of the highest position in the hopper. After the unloading begins, the gate is initially fully open.

[0067] At the start of loading, the belt conveyor is turned on and continuously feeds material into one of the three bins. For simplicity, this embodiment will not describe the feeding process, but will use "record the weight value of the highest material in the bin" to indicate that the bin to be unloaded has accumulated more material than the target weight value for unloading from that bin.

[0068] Step 202, Material release weight monitoring: Weight monitoring is performed in real time through the weighing sensors in the hopper.

[0069] Before the unloading gate of the hopper is opened, the weight of the material displayed by the load cell is the actual weight of the material in the hopper. This weight should be slightly greater than the target weight value, which is the aforementioned "record the weight value of the highest point of the material in the hopper". After the unloading gate is opened, the load cell continuously records the decrease in weight in the hopper during the unloading process, and then cross-checks the loading weight value with the flow rate value.

[0070] Step 203, Material discharge flow control: The flow rate is monitored in real time by detecting the gate opening, the amount of material flowing out, and the changes in the accumulation of material in the carriage.

[0071] PID monitors the gate opening process using the following formula:

[0072]

[0073] The size of the dead zone and the lead time are given based on the learning results of the learning machine.

[0074] As shown in the above formula, the controlled position does not ultimately need to be completely closed. The action stops after reaching a certain position to ensure the material falls. The gate is then fully closed only when the target value is reached after subtracting a lead time from the falling amount. The purpose of the lead time is to ensure that some material will continue to fall even at the moment the gate is fully closed. By estimating this lead time, the amount of material released after the gate is closed is closer to the target value.

[0075] Step 204, End of material feeding: The controller obtains the current dead zone and lead time of the loading based on the learning and training. The gate stops at the lead time position, and the material continues to be fed down slowly until the target value is reached. The gate then closes completely, and the feeding ends.

[0076] Each action will be stored as experience to help with future actions.

[0077] Finally, it should be noted that the above is only used to illustrate the technical solution of the present invention and not to limit it. Although the present invention has been described in detail with reference to the preferred arrangement, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solution of the present invention (such as the form of the loading station, the application of various formulas, the order of steps, etc.) without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. A method for adaptive high-precision control of material feeding, wherein the loading station used in the method is a multi-bucket rotary loading system, characterized in that, The method includes a learning and training process and a normal material feeding process: The learning and training process includes the following steps: Step 101, Data Collection: Collect information on the discharge of various bulk materials in the past, including the flowability of the materials, the size of the dead zone, the amount of lead time, and the parameters of gate closure; Step 102, manual operation: using materials with various characteristics to carry out actual manual operation of feeding materials, collecting various gate parameters during the operation process, including material flowability and dead zone size, lead time, and gate closing parameters; The fluidity is determined based on the monitoring of material movement and material accumulation by sensors during actual operation. Step 103, Learning and Training: Input the collected data and manual operation data into the database, and train the gate to close using a learning machine, so as to achieve adaptive gate closing to dead zone and lead time; the database will be continuously improved in the subsequent normal material feeding process; The normal material feeding process includes the following steps: Step 201, Open the gate to release material: After receiving the release command, first record the weight value of the highest position in the hopper. After the unloading begins, the gate is initially fully open. Step 202, Material release weight monitoring: Weight monitoring is performed in real time through the weighing sensors in the hopper; Step 203, Material discharge flow control: The flow rate is monitored in real time by the gate opening, the amount of material flowing out, and the changes in the accumulation of material in the carriage; PID monitors the gate opening process using the following formula: Among them, the size of the dead zone and the size of the lead are given based on the learning results of the learning machine; Step 204, End of material feeding: The controller obtains the current dead zone and lead time of the loading based on the learning and training. The gate stops at the lead time position, and the material continues to be fed down slowly until the target value is reached. The gate then closes completely, and the feeding ends.

Images