A Method and System for Separating and Tracing Magnetic Products Based on Blockchain and Multimodal Sensing

Abstract

This invention proposes a method and system for separating and tracing magnetic products based on blockchain and multimodal sensing. The method includes: acquiring magnetic induction signals through a sensor array; calculating the adsorption coefficient based on the product's physical parameters and the magnetic induction signals, and outputting the stacking state level; outputting the corresponding cylinder thrust and stroke based on the adsorption coefficient and stacking state level to perform a separation operation; adjusting the cylinder thrust and stroke during actual execution through an error correction mechanism; calculating the final spacing after product stabilization based on the adjusted cylinder thrust and stroke; collecting the thrust, stroke, and spacing data for each separation operation in the aforementioned steps, and generating a hash value that is bound to the product tag. This invention, by introducing a magnetic state sensing and control mechanism, enables the separation device to automatically adapt the separation path and thrust matching according to the degree of adsorption and stacking state between products, achieving rapid determination and precise fine-tuning of the separation result.

Description

Technical Field

[0001] This invention belongs to the field of intelligent manufacturing, and in particular relates to a method and system for separating and tracing magnetic products based on blockchain and multimodal sensing. Background Technology

[0002] In current automated packaging and sorting production scenarios for magnetic products, products are typically stacked continuously and transported to the next stage via a conveyor. Due to their inherent magnetism, these products are prone to mutual attraction during stacking, forming tight connections that pose significant challenges to subsequent separation, picking, and positioning. To address these issues, existing separation methods generally employ a cylinder-driven mechanism combined with simple sensing systems such as photoelectric sensors or through-beam optical fibers to mechanically separate the products. However, existing technologies suffer from several problems: Firstly, existing equipment often uses fixed-cycle drives or static parameter settings, failing to flexibly adjust the separation action based on differences in adsorption strength, stacking morphology, or product model. This leads to problems such as material clamping, material sticking, emptying, or product surface damage in practical applications. Secondly, existing structures rely primarily on mechanical separation methods, resulting in low levels of intelligence and limited feedback, failing to accurately sense dynamic adsorption states and provide flexible responses. Furthermore, due to the limitations of mechanical structure adaptability and sensor detection capabilities, changing product models or switching between different batches of magnetic products often requires downtime for structural adjustments or resetting control parameters, significantly impacting production line efficiency and flexibility. In summary, while existing magnetic product separation devices can play a role in certain scenarios, they still have shortcomings in terms of accuracy, adaptability, automation level, and stability. In particular, they cannot meet the core requirements of modern multi-category, high-paced intelligent production lines for "flexible separation" and "non-destructive processing." Summary of the Invention

[0003] The purpose of this invention is to design a method and system for separating and tracing magnetic products based on blockchain and multimodal sensing, which can provide a solution for separating magnetic products with rapid changeover, and improve the overall operating efficiency and product processing quality of the system.

[0004] To achieve the above objectives, a method for separating and tracing magnetic products based on blockchain and multimodal sensing is provided in a first aspect of the present invention, the method comprising:

[0005] Magnetic induction signals are collected by a sensor array, the adsorption coefficient is calculated based on the product's physical parameters and the magnetic induction signals, and the stacking state level is output; the stacking state level is divided into three levels according to a preset first threshold and a second threshold.

[0006] Separation operations are performed based on the cylinder thrust and stroke corresponding to the adsorption coefficient and stacking state level. The cylinder thrust is obtained by combining the adsorption coefficient and stacking state level and introducing a regularization term to smoothly control the thrust increase. The stroke is calculated using the adsorption coefficient, stacking state level, and adjustment function. The adjustment function adjusts the influence of the stacking state level on the stroke through a smoothing function. When the stacking state level is single-piece, the adjustment function does not affect the stroke. When it is at other stacking state levels, the adjustment function increases the stroke.

[0007] An error correction mechanism is used to adjust the cylinder thrust and stroke during actual execution. The error correction mechanism calculates error terms in real time and adjusts the thrust and stroke based on the error terms. The error terms represent the difference between the actual position and the target position of the cylinder.

[0008] The final spacing after product stabilization is calculated based on the adjusted cylinder thrust and stroke; the final spacing is obtained by adding the calculated actual spacing to the stroke error; the actual spacing is obtained by weighted calculation of the preset target spacing and the adjusted cylinder thrust and stroke.

[0009] The thrust, stroke, and spacing data of each separation operation are collected to obtain operation records. A hash value is generated for each operation record and bound to the corresponding product label.

[0010] Furthermore, the adsorption coefficient is obtained by multiplying the average value of the local magnetic induction intensity deviation as the adsorption field strength with the physical parameters of the product; the physical parameters of the product include the material's magnetic permeability and thickness.

[0011] Furthermore, the stacking status level classification specifically includes: when the adsorption coefficient is lower than the first threshold, the level value is 0, and the stacking status is a single product; when the adsorption coefficient is higher than the first threshold and lower than the second threshold, the level value is 1, and the stacking status is two products stacked; when the adsorption coefficient is higher than the second threshold, the level value is 2, and the stacking status is three or more products stacked.

[0012] Furthermore, the regularization term keeps the thrust stable when the adsorption coefficient is less than a preset first threshold, and increases the thrust when the adsorption coefficient is higher than the preset first threshold.

[0013] Furthermore, the error correction mechanism is based on a PID control algorithm to calculate and correct the deviation of cylinder thrust and stroke in real time; the proportional, integral and derivative coefficients of the PID control algorithm are used to adjust the thrust response, accumulate error and suppress high-frequency oscillation, respectively.

[0014] Furthermore, the error correction mechanism also introduces a delay compensation term, which compensates for the delay in advance by adjusting the thrust. The delay is obtained through real-time feedback measurement from sensors.

[0015] Furthermore, the actual spacing is adjusted based on the change in the adsorption coefficient to avoid the product being re-adsorbed due to an excessively strong adsorption coefficient.

[0016] Furthermore, the travel error is the difference between the adjusted travel and the preset target travel.

[0017] Furthermore, the hash value includes a unique identifier for the product batch, a unique identifier for the operator, the final spacing, and the adjusted cylinder thrust and stroke.

[0018] A second aspect of the invention provides a magnetic product separation and traceability system based on blockchain and multimodal sensing, the system comprising:

[0019] The status detection unit is used to collect magnetic induction signals through a sensor array, calculate the adsorption coefficient based on the product's physical parameters and the magnetic induction signals, and output the stacking status level; the stacking status level is divided into three levels according to a preset first threshold and a second threshold.

[0020] The motion planning unit is used to output the corresponding cylinder thrust and stroke based on the adsorption coefficient and stacking state level for separation operation. The cylinder thrust is obtained by combining the adsorption coefficient and stacking state level and introducing a regularization term to smoothly control the thrust increase. The stroke is calculated by the adsorption coefficient, stacking state level and adjustment function. The adjustment function adjusts the influence of stacking state level on stroke through a smoothing function. When the stacking state level is single piece, the adjustment function has no effect on the stroke. When it is in other stacking state levels, the adjustment function increases the stroke.

[0021] The motion calibration unit is used to adjust the cylinder thrust and stroke during actual execution through an error correction mechanism; the error correction mechanism calculates error terms in real time and adjusts the thrust and stroke based on the error terms; the error terms represent the difference between the actual position and the target position of the cylinder;

[0022] The separation confirmation unit is used to calculate the final spacing after product stabilization by adjusting the cylinder thrust and stroke; the final spacing is obtained by adding the calculated actual spacing to the stroke error; the actual spacing is obtained by weighted calculation of the preset target spacing and the adjusted cylinder thrust and stroke.

[0023] The data collection unit is used to collect thrust, stroke, and spacing data for each separation operation to obtain operation records, and to generate a hash value for each operation record and bind it to the corresponding product label.

[0024] The beneficial technical effects of the present invention are at least as follows:

[0025] To address the aforementioned issues, this invention provides a method and system for separating and tracing magnetic products based on blockchain and multimodal sensing. By introducing a magnetic state sensing mechanism and a flexible control strategy, the separation device can automatically adapt the separation path and thrust according to the degree of adsorption and stacking state between products. Simultaneously, a dynamic correction unit with end-point identification capability is incorporated to achieve rapid determination and precise fine-tuning of the separation results, significantly improving the system's ability to cope with different stacking methods and product specifications. Furthermore, this invention employs a modular design, facilitating rapid switching between different models of magnetic products and migration of control strategies, thereby reducing manual intervention and adjustment time and improving the continuous operation efficiency of the production line. Attached Figure Description

[0026] The present invention will be further described with reference to the accompanying drawings, but the embodiments in the drawings do not constitute any limitation on the present invention. For those skilled in the art, other drawings can be obtained based on the following drawings without creative effort.

[0027] Figure 1 This is a flowchart of a magnetic product separation and traceability method based on blockchain and multimodal sensing according to the present invention.

[0028] Figure 2 This is a framework diagram of a magnetic product separation and traceability system based on blockchain and multimodal sensing according to the present invention. Detailed Implementation

[0029] Embodiments of the present invention are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0030] In one or more embodiments, such as Figure 1 As shown, a method for separating and tracing magnetic products based on blockchain and multimodal sensing is disclosed. The method includes the following:

[0031] S1: Collect magnetic induction signals through a sensor array, calculate the adsorption coefficient based on the product's physical parameters and the magnetic induction signals, and output the stacking state level; the stacking state level is divided into three levels according to a preset first threshold and a second threshold;

[0032] Specifically, in this embodiment, the core objective of this step is to estimate the adsorption coefficient between products and determine the product stacking state by acquiring magnetic field signals in real time. This information will directly affect the thrust setting and stroke control of the subsequent cylinder control. Magnetic induction signals are acquired through a Hall sensor array, and combined with the physical properties of the products, the adsorption coefficient is quantitatively modeled.

[0033] First, this entire process relies on real-time data acquisition from a Hall sensor array. The primary function of the Hall sensor is to calculate magnetic field strength by measuring the change in the magnetic field generated when an object passes by. In this embodiment, multiple Hall sensors are installed above or to the side of the separation area to detect the magnetic adsorption state between the products. Each sensor samples every 20 milliseconds, outputting a voltage signal. It will be through calibration coefficients Convert to magnetic induction intensity value Calibration coefficient It was obtained after experimental calibration based on the sensor's sensitivity.

[0034] In this embodiment, if the voltage output by the measured Hall sensor is , After conversion, the corresponding magnetic flux density value can be obtained. .

[0035] In this embodiment, the input data for this step comes from two parts:

[0036] First, a Hall effect magnetic field sensor array is installed in a row above or to the side of the product feed channel. There are (e.g., 6) Hall effect sensors. These sensors sample once every 20ms and output voltage values. (Unit: mV), the system internally converts it into magnetic field strength using a preset calibration relationship. (Unit is mT), that is ,in The sensitivity coefficients of each sensor (e.g., 2.5 mT / mV) are factory-calibrated and imported into the controller during the system initialization phase.

[0037] Second, the product's basic physical parameters are either manually set by the system during model switching or retrieved from the process database. These parameters include the thickness of the current product model. (e.g., 1.8 mm), approximate value of material permeability (For example) ), and recommended safety clearances between individual products (e.g., 2.5mm). This information is stored in the controller as a "static reference value" and is not involved in runtime learning or adjustment.

[0038] During the operation of the conveyor line, the Hall sensor array detects localized changes in magnetic induction intensity whenever a product passes through the magnetic sensing zone. For a single product, the magnetic field strength exhibits a gentle peak shape in space; however, if two or more products are attracted together, the magnetic field will show strong peak changes or abrupt shifts. By detecting multiple... By analyzing the spatial variation of the value and combining it with material and structural information, the adsorption coefficient can be estimated.

[0039] First, check all sensor output values. Perform a linear fit and calculate the average deviation. Based on this, an adsorption coefficient estimation formula is constructed:

[0040]

[0041] This formula utilizes the average value of local magnetic induction intensity deviation as the spatial characteristic of the adsorption field strength, and then multiplies it by the material permeability of this product model. and thickness This allows for the estimation of the stacking adsorption coefficient, with all sensor outputs and other relevant parameters having been appropriately normalized.

[0042] In this embodiment, a 1.8mm thick magnetic blade model (steel, with an approximate magnetic permeability) was tested. When the two blades are attracted, the adjacent Hall sensors The value can vary by ±20mT. After inputting the above model, the adsorption coefficient is estimated to be 0.85, and the system records it as "medium adsorption strength". The medium thrust parameter will be set in subsequent actions.

[0043] Furthermore, in order to determine whether multiple products are adsorbed, a structured level judgment logic is also needed. This is used to determine whether the current adsorption is for a single item, two items, or three or more items. The definition is as follows:

[0044]

[0045] in, The adsorption coefficient is . and The threshold value is calibrated in actual testing, and the unit is N. Based on experimental data, in this embodiment, the threshold value is set as follows: and , representing the adsorption limit and the hazardous adsorption limit, respectively. This structure is a simple and fast discrete classification method that does not require real-time classifier modeling, facilitating deployment in industrial edge controllers.

[0046] S2: Separation operation is performed based on the cylinder thrust and stroke corresponding to the adsorption coefficient and stacking state level. The cylinder thrust is obtained by combining the adsorption coefficient and stacking state level and introducing a regularization term to smoothly control the thrust increase. The stroke is calculated by the adsorption coefficient, stacking state level and adjustment function. The adjustment function adjusts the influence of stacking state level on stroke through a smoothing function. When the stacking state level is single piece, the adjustment function does not affect the stroke. When it is in other stacking state levels, the adjustment function increases the stroke.

[0047] Specifically, in this embodiment, the core objective of this step is to apply the adsorption coefficient estimated in the previous step. and stacking state This involves planning control parameters for the separation operation. This is achieved by controlling the thrust of the cylinder. ,journey and termination position This ensures accurate separation of magnetic products with varying adsorption strengths and stacking configurations without damage. A flexible control algorithm is introduced in this step, dynamically adjusting thrust and stroke based on real-time feedback to adapt to separation requirements under different stacking conditions. Based on the adsorption coefficient... and stacking state In this embodiment, the thrust and stroke control parameters are dynamically generated using the following algorithm:

[0048] Thrust generation: Thrust This is the main control variable of the cylinder during the separation process. Its magnitude needs to be flexibly adjusted according to the current adsorption coefficient and stacking state. Especially in the case of multi-piece adsorption, the range of thrust variation needs to be increased to overcome the large adsorption coefficient and ensure smooth separation.

[0049] To ensure that the thrust during the separation process can flexibly adapt to different adsorption coefficients and stacking states, this step involves precise calculation of the thrust. This allows the separation process to automatically adjust according to actual conditions. Especially when adsorbing multiple products, the thrust needs to be dynamically increased to overcome stronger adsorption forces and ensure smooth separation. The thrust calculation formula incorporates the adsorption coefficient... and stacking state Taking all factors into consideration, a regularization term is introduced to smoothly control the increase in thrust, preventing excessive thrust from causing equipment overload or product damage when the adsorption coefficient is high. This refined thrust generation method allows for dynamic adjustment of the thrust during the separation process based on different adsorption conditions and stacking states, ensuring stable product separation. The thrust calculation formula is as follows:

[0050]

[0051] in: The cylinder thrust is dynamically adjusted over time, directly controlling the intensity of the separation process; The adsorption coefficient calculated from the previous step determines the baseline magnitude of the thrust; The stacked state affects the increase in thrust required for separation; This is the experimental calibration coefficient, used to adjust the relationship between the magnitude of the thrust and the adsorption coefficient and the stacking state; This is an adjustment coefficient used to control the smoothness of the thrust response; The threshold value for the adsorption coefficient determines how the thrust changes when the adsorption coefficient exceeds a certain critical value. The introduced regularization term serves to: when the adsorption coefficient is less than... When the adsorption coefficient is large, the thrust remains stable, while when the adsorption coefficient is large, the thrust increases rapidly, thus avoiding the impact on the system when the thrust is too high.

[0052] Furthermore, the cylinder stroke varies depending on the adsorption coefficient and stacking state. Fine-tuning is required to ensure the smooth completion of the separation process. For higher adsorption coefficients, the cylinder needs a longer stroke to completely separate the product. To ensure the smooth completion of the separation process under different adsorption coefficients and stacking conditions, this step involves precisely calculating the cylinder stroke. This allows each separation action to be adapted to the specific characteristics of the product. Especially in cases of high adsorption coefficients and multiple stacked items, the cylinder stroke needs to be increased to ensure complete product separation and prevent jamming. The stroke adjustment is not only based on the adsorption coefficient... It also considers stacking states. By dynamically optimizing the stroke through a smooth adjustment function, unnecessary stroke waste is avoided when the stacking is small, ensuring the efficiency and accuracy of the separation process.

[0053] The formula for calculating the distance is as follows:

[0054]

[0055] in: The cylinder stroke determines the distance the cylinder travels during product separation. The adsorption coefficient directly affects the required stroke of the cylinder; The stacked state affects the size of the stroke; These are experimental calibration coefficients used to adjust the relationship between stroke, adsorption coefficient, and stacking state; The adjustment coefficient determines the impact of the stacking state on the stroke; As an adjustment function, when the stack is a single item, the stroke is kept small to avoid wasting space; when the stack is multiple items, the stroke is increased to ensure complete separation.

[0056] The key to this formula lies in the stroke adjustment function, which adjusts the impact of stacking state on stroke through a smoothing function, thereby optimizing the stroke of the cylinder under different stacking states. This avoids the excessive stroke caused by excessive stacking in traditional separation schemes, and maintains separation efficiency and accuracy.

[0057] S3: The cylinder thrust and stroke during actual execution are adjusted through an error correction mechanism; the error correction mechanism calculates error terms in real time and adjusts the thrust and stroke based on the error terms; the error terms represent the difference between the actual position and the target position of the cylinder;

[0058] Specifically, in this embodiment, this step introduces a real-time error feedback mechanism and an adaptive control algorithm to finely adjust the cylinder's trajectory based on the actual execution of thrust and stroke. This process not only ensures smooth product separation but also addresses system errors and improves overall separation accuracy, especially in complex situations such as multi-piece adsorption and different stacking states.

[0059] In this step, the thrust and stroke of the cylinder need to be finely adjusted to ensure that the actual motion trajectory is consistent with the set target. To this end, this solution designs a real-time feedback error correction mechanism, which monitors the actual position and thrust of the cylinder through sensor feedback data, thereby correcting the thrust and stroke and eliminating errors caused by equipment lag, friction, etc.

[0060] An error compensation mechanism based on a PID control algorithm is employed here to calculate and correct deviations in cylinder position and thrust in real time. The PID control algorithm achieves high-precision control through dynamic feedback adjustment of thrust and stroke errors.

[0061] First, introduce the error term. It represents the difference between the actual position and the target position of the cylinder:

[0062]

[0063] in: This is the error term, representing the actual position of the cylinder. With the target location The differences between them. It is the target position set according to the separation action, and is usually calculated in the second step. The actual cylinder position is obtained through sensor measurement.

[0064] To further improve the accuracy and stability of the separation process, this embodiment introduces a PID control algorithm to dynamically adjust the cylinder's thrust and stroke. Since the actual position of the cylinder may be affected by external factors (such as friction, delayed response, etc.) during the separation process, it is necessary to monitor and correct the error in real time to ensure that the cylinder accurately reaches the target position. Through the PID controller, this solution can automatically adjust the thrust and stroke according to the magnitude and trend of the error to minimize the error and avoid system oscillation. This embodiment uses a PID control formula to adjust the cylinder's thrust and stroke to reduce error. The formula is as follows:

[0065]

[0066] in: It is the thrust parameter after error correction, which is actually used for cylinder control. These are the preliminary thrust parameters obtained from the second step. , and These are the proportional, integral, and derivative coefficients of the PID controller, which are used to adjust the thrust response, accumulate error, and suppress high-frequency oscillations, respectively. It is the position error, which calculates the difference between the actual position of the cylinder and the target position.

[0067] The key to this formula lies in its real-time adjustment of feedback error, which enables the system to respond quickly and correct deviations, avoiding the long-term error accumulation caused by lag or equipment differences in traditional control schemes. It is especially suitable for high-precision separation requirements.

[0068] Furthermore, this embodiment also introduces a delay compensation term because the cylinder response in the system may have a time lag, especially under high load or low speed. To reduce the impact of this delay on separation accuracy, this scheme compensates for the delay in advance by adjusting the thrust, as shown in the following formula:

[0069]

[0070] in: This is the adjusted thrust. It is the delay compensation coefficient, which is obtained through experimental calibration. It is the delay time of cylinder action, which is usually measured in real time by sensor feedback.

[0071] The introduction of this compensation item ensures that even when the system response is slow, the cylinder can adjust in advance, thereby reducing errors caused by delay and ensuring the precise execution of the separation action.

[0072] S4: The final spacing after product stabilization is calculated based on the adjusted cylinder thrust and stroke; the final spacing is obtained by adding the calculated actual spacing to the stroke error; the actual spacing is obtained by weighted calculation of the preset target spacing and the adjusted cylinder thrust and stroke.

[0073] Specifically, in practical applications, especially in high-speed production environments, the distance between separated products may deviate slightly due to factors such as equipment response delays, friction, and external disturbances. To ensure the smooth operation of subsequent automated processes such as robotic gripping and packaging, the distance between products must be accurately confirmed and maintained stable. This process cannot rely solely on a pre-set target distance but requires dynamic adjustment of the actual distance after separation.

[0074] After the separation operation is completed, to ensure a stable spacing between the products, this solution designs a dynamic spacing maintenance model based on thrust and stroke. This model combines factors such as thrust, stroke, and adsorption coefficient, and through precise control and adjustment, ensures that the finally separated products maintain an appropriate and stable spacing. In practical applications, the adsorption force and stacking state of the products may change; therefore, this model has real-time adjustment capabilities, flexibly adjusting the spacing according to these changes to ensure that subsequent automated processing is not disturbed. The dynamic spacing maintenance model based on thrust and stroke is designed here, and the final separation spacing is adjusted using the following formula:

[0075]

[0076] in: Indicates time The actual separation distance between products at any given time, expressed in spacing units. The target spacing is the preset standard value (such as 2.5mm) based on the product type and production requirements; it is the preset initial spacing. The thrust obtained through the fine-tuning control in the third step is used to control the mechanical output of the cylinder. This is the stroke after error correction, used to control the accurate movement of the cylinder. These are coefficients obtained through experimental calibration, used to adjust the specific effects of thrust and stroke on the spacing. The adsorption coefficient estimate from step S1 represents the magnetic adsorption coefficient between products. This is the threshold value for the adsorption coefficient, indicating that when the adsorption coefficient is greater than this value, the product spacing needs to be adjusted further. As a regularization term, the spacing is dynamically adjusted according to the change of adsorption coefficient. When the adsorption coefficient is large, the system will automatically increase the separation spacing to avoid the product being re-adsorbed due to an excessively strong adsorption coefficient.

[0077] Furthermore, to precisely adjust the separation distance, a stroke fine-tuning compensation mechanism is introduced to address potential stroke errors. The actual separation stroke may deviate slightly due to cylinder friction, delay, or other factors. The final separation distance is fine-tuned using the following formula:

[0078]

[0079] in: The final adjusted separation distance represents the stable product spacing after compensation. The target stroke is the ideal stroke set according to the requirements of the separation process. The coefficients obtained through experiments are used to control the impact of stroke error on the final pitch. This compensation mechanism is particularly suitable for small errors that may exist in the equipment, ensuring that the final pitch can be precisely controlled even under the influence of cylinder response delay or friction, and avoiding error accumulation caused by unstable equipment performance.

[0080] S5: Collect thrust, stroke, and spacing data for each separation operation to obtain operation records, generate a hash value for each operation record, and bind it to the corresponding product label.

[0081] Specifically, the goal of this step is to ensure that each operation is linked to specific production batches, production line status, and operators by collecting and storing key data during the separation process, forming a complete production data chain to ensure subsequent quality control and traceability of the production process. Through smart tags and a database system, data is tightly associated with each product or batch, supporting product quality traceability. The entire solution relies on precise data storage and tamper-proof technology to ensure data integrity and immutability. To ensure that data from each separation operation is accurately stored and linked to specific information in the production process (such as production batches and operators), a data storage scheme based on database records and blockchain technology is designed here.

[0082] Data such as thrust, stroke, and spacing for each separation operation are collected in real time by sensors in the system and recorded along with relevant information for that operation (such as production batch and operator information). To ensure data security and tamper resistance, this step generates a unique hash value for each record and binds it to the product's smart tag. The hash calculation formula for the data record is as follows:

[0083]

[0084] in: This represents the hash value of the data being operated on, and is a unique identifier generated. The thrust parameters are the fine-tuned parameters from step S4. These are the fine-tuned stroke parameters from step S4. This is the final separation distance in step S4. A unique identifier for a product batch. This hash value serves as a unique identifier for operators, used to track responsibility and operational status. It ensures that data from each separation operation is unique and tamper-proof, providing precise evidence for product quality traceability.

[0085] Each time data is generated, it is immediately recorded in the blockchain and linked to the product's smart tag. The smart tag, acting as a unique identifier for the product, records information about the entire process of that batch of products from production to distribution, including operational data such as thrust, stroke, and separation distance for each step. This data is permanently stored in the blockchain, ensuring its immutability and traceability. Each product or batch's smart tag corresponds to a hash value recorded in the blockchain, enabling full traceability of product information.

[0086] In one or more embodiments, such as Figure 2 As shown, a magnetic product separation and traceability system based on blockchain and multimodal sensing is disclosed. The system includes:

[0087] The status detection unit is used to collect magnetic induction signals through a sensor array, calculate the adsorption coefficient based on the product's physical parameters and the magnetic induction signals, and output the stacking status level; the stacking status level is divided into three levels according to a preset first threshold and a second threshold.

[0088] The motion planning unit is used to output the corresponding cylinder thrust and stroke based on the adsorption coefficient and stacking state level for separation operation. The cylinder thrust is obtained by combining the adsorption coefficient and stacking state level and introducing a regularization term to smoothly control the thrust increase. The stroke is calculated by the adsorption coefficient, stacking state level and adjustment function. The adjustment function adjusts the influence of stacking state level on stroke through a smoothing function. When the stacking state level is single piece, the adjustment function has no effect on the stroke. When it is in other stacking state levels, the adjustment function increases the stroke.

[0089] The motion calibration unit is used to adjust the cylinder thrust and stroke during actual execution through an error correction mechanism; the error correction mechanism calculates error terms in real time and adjusts the thrust and stroke based on the error terms; the error terms represent the difference between the actual position and the target position of the cylinder;

[0090] The separation confirmation unit is used to calculate the final spacing after product stabilization by adjusting the cylinder thrust and stroke; the final spacing is obtained by adding the calculated actual spacing to the stroke error; the actual spacing is obtained by weighted calculation of the preset target spacing and the adjusted cylinder thrust and stroke.

[0091] The data collection unit is used to collect thrust, stroke, and spacing data for each separation operation to obtain operation records, and to generate a hash value for each operation record and bind it to the corresponding product label.

[0092] It is worth noting that the specific workflow of the magnetic product separation and traceability system based on blockchain and multimodal sensing provided in this embodiment of the invention is the same as that of the magnetic product separation and traceability method based on blockchain and multimodal sensing described in the above embodiments, and will not be repeated here.

[0093] This invention also provides a magnetic product separation and traceability device based on blockchain and multimodal sensing, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the steps described in the above embodiment of a magnetic product separation and traceability method based on blockchain and multimodal sensing. Figure 1 The steps S1 to S5 described above; or, when the processor executes the computer program, it implements the functions of each module in the above system embodiments.

[0094] For example, the computer program may be divided into one or more modules, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules may be a series of computer program instruction segments capable of performing specific functions, which describe the execution process of the computer program in the magnetic product separation and traceability device based on blockchain and multimodal sensing.

[0095] The magnetic product separation and traceability device based on blockchain and multimodal sensing can be a computing device such as a desktop computer, laptop, handheld computer, or cloud server. This device may include, but is not limited to, processors and memory. Those skilled in the art will understand that the device may also include input / output devices, network access devices, buses, etc.

[0096] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor. This processor is the control center of the blockchain-based and multimodal sensing magnetic product separation and traceability device, connecting all parts of the device via various interfaces and lines.

[0097] The memory can be used to store the computer program and / or modules. The processor, by running or executing the computer program and / or modules stored in the memory, and by calling the data stored in the memory, realizes various functions of the magnetic product separation and traceability device based on blockchain and multimodal sensing. The memory may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function, etc.; the data storage area may store data created based on the operation of the air conditioner controller, etc. In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart memory card (SMC), secure digital card (SD), flash memory card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0098] The module integrated with the blockchain and multimodal sensing-based magnetic product separation and traceability device, if implemented as a software functional unit and sold or used as an independent product, can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0099] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0100] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for separating and tracing the origin of magnetic products based on blockchain and multimodal sensing, characterized in that, The method includes: Magnetic induction signals are collected by a sensor array, the adsorption coefficient is calculated based on the product's physical parameters and the magnetic induction signals, and the stacking state level is output; the stacking state level is divided into three levels according to a preset first threshold and a second threshold. Separation operations are performed based on the cylinder thrust and stroke corresponding to the adsorption coefficient and stacking state level. The cylinder thrust is obtained by combining the adsorption coefficient and stacking state level and introducing a regularization term to smoothly control the thrust increase. The stroke is calculated using the adsorption coefficient, stacking state level, and adjustment function. The adjustment function adjusts the influence of the stacking state level on the stroke through a smoothing function. When the stacking state level is single-piece, the adjustment function does not affect the stroke. When it is at other stacking state levels, the adjustment function increases the stroke. An error correction mechanism is used to adjust the cylinder thrust and stroke during actual execution. The error correction mechanism calculates error terms in real time and adjusts the thrust and stroke based on the error terms. The error terms represent the difference between the actual position and the target position of the cylinder. The final spacing after product stabilization is calculated based on the adjusted cylinder thrust and stroke; the final spacing is obtained by adding the calculated actual spacing to the stroke error; the actual spacing is obtained by weighted calculation of the preset target spacing and the adjusted cylinder thrust and stroke. The thrust, stroke, and spacing data of each separation operation are collected to obtain operation records. A hash value is generated for each operation record and bound to the corresponding product label.

2. The method for separating and tracing magnetic products based on blockchain and multimodal sensing according to claim 1, characterized in that, The adsorption coefficient is obtained by multiplying the average value of the local magnetic induction intensity deviation as the adsorption field strength with the physical parameters of the product; the physical parameters of the product include the material's magnetic permeability and thickness.

3. The method for separating and tracing magnetic products based on blockchain and multimodal sensing according to claim 1, characterized in that, The stacking status classification specifically includes: when the adsorption coefficient is lower than the first threshold, the level is 0, and the stacking status is a single product; when the adsorption coefficient is higher than the first threshold and lower than the second threshold, the level is 1, and the stacking status is two products stacked; when the adsorption coefficient is higher than the second threshold, the level is 2, and the stacking status is three or more products stacked.

4. The method for separating and tracing magnetic products based on blockchain and multimodal sensing according to claim 1, characterized in that, The regularization term keeps the thrust stable when the adsorption coefficient is less than a preset first threshold, and increases the thrust when the adsorption coefficient is higher than the preset first threshold.

5. The method for separating and tracing magnetic products based on blockchain and multimodal sensing according to claim 1, characterized in that, The error correction mechanism is based on a PID control algorithm to calculate and correct the deviation of cylinder thrust and stroke in real time. The proportional, integral and derivative coefficients of the PID control algorithm are used to adjust the thrust response, accumulate error and suppress high-frequency oscillation, respectively.

6. The method for separating and tracing magnetic products based on blockchain and multimodal sensing according to claim 5, characterized in that, The error correction mechanism also introduces a delay compensation term, which compensates for the delay in advance by adjusting the thrust. The delay is obtained by real-time feedback measurement from the sensor.

7. The method for separating and tracing magnetic products based on blockchain and multimodal sensing according to claim 1, characterized in that, The actual spacing is also adjusted according to the change of the adsorption coefficient to avoid the product being re-adsorbed due to an excessively strong adsorption coefficient.

8. The method for separating and tracing magnetic products based on blockchain and multimodal sensing according to claim 1, characterized in that, The travel error is the difference between the adjusted travel and the preset target travel.

9. The method for separating and tracing magnetic products based on blockchain and multimodal sensing according to claim 1, characterized in that, The hash value includes a unique identifier for the product batch, a unique identifier for the operator, the final spacing, and the adjusted cylinder thrust and stroke.

10. A magnetic product separation and traceability system based on blockchain and multimodal sensing, characterized in that, The system includes: The status detection unit is used to collect magnetic induction signals through a sensor array, calculate the adsorption coefficient based on the product's physical parameters and the magnetic induction signals, and output the stacking status level; the stacking status level is divided into three levels according to a preset first threshold and a second threshold. The motion planning unit is used to output the corresponding cylinder thrust and stroke based on the adsorption coefficient and stacking state level for separation operation. The cylinder thrust is obtained by combining the adsorption coefficient and stacking state level and introducing a regularization term to smoothly control the thrust increase. The stroke is calculated by the adsorption coefficient, stacking state level and adjustment function. The adjustment function adjusts the influence of stacking state level on stroke through a smoothing function. When the stacking state level is single piece, the adjustment function has no effect on the stroke. When it is in other stacking state levels, the adjustment function increases the stroke. The motion calibration unit is used to adjust the cylinder thrust and stroke during actual execution through an error correction mechanism; the error correction mechanism calculates error terms in real time and adjusts the thrust and stroke based on the error terms; the error terms represent the difference between the actual position and the target position of the cylinder. The separation confirmation unit is used to calculate the final spacing after product stabilization by adjusting the cylinder thrust and stroke; the final spacing is obtained by adding the calculated actual spacing to the stroke error; the actual spacing is obtained by weighted calculation of the preset target spacing and the adjusted cylinder thrust and stroke. The data collection unit is used to collect thrust, stroke, and spacing data for each separation operation and generate hash values ​​that are then linked to product labels.

Images