Embedded terminal remote configuration method and device based on Internet of Things, and medium

Abstract

The invention relates to the technical field of digital information transmission, and discloses an embedded terminal remote configuration method and device based on the Internet of Things and a medium, and the method comprises the steps that a target embedded terminal receives a configuration parameter set and pre-stores the configuration parameter set in a preliminary buffer area, and the configuration parameter set is in a to-be-activated state; acquiring a background signal sequence in a working condition no-load intermittent period and determining a dynamic trigger threshold value; extracting an energy feature of the physical event trigger signal flow to generate a feature vector; by comparing the feature vector with a dynamic trigger threshold value, locking a feature vector take-off point to identify a synchronous trigger moment; in response to the synchronous triggering moment, the loading instruction is executed to cover the configuration parameter set to the running register, a transient signal generated by a physical event is used as a loading triggering source, so that configuration switching is converted from communication protocol driving to physical field direct triggering, the problem of parameter effectiveness uncertainty caused by transmission delay fluctuation is solved, and the reliability of configuration switching is improved. And accurate alignment of the configuration effective node and the physical production rhythm is ensured.

Description

Technical Field

[0001] This invention belongs to the field of digital information transmission technology, and in particular relates to a method, device and medium for remote configuration of embedded terminals based on the Internet of Things. Background Technology

[0002] Remote configuration of embedded terminals is a key link in realizing flexible production. By using gateways to distribute configuration parameter sets for different production processes to each sensor node, in production scenarios involving high-speed physical deformation such as continuous metal rolling, the signal processing parameters of the sensor nodes need to be switched transiently at the millisecond or even microsecond level according to the physical characteristics of the workpiece to be rolled, so as to ensure that the data acquisition features are aligned with the processing cycle.

[0003] Industrial wireless networks or time-sensitive networks generally suffer from unavoidable link jitter during data transmission, resulting in uncertainty in the arrival time of configuration commands at the terminal. If a command parsing-driven mode is directly adopted, the configuration takes effect after the moment the metal enters the roll. This mismatch between information flow and physical flow causes mismatch in sampling range or filter cutoff frequency during periods of drastic changes in operating conditions, leading to instantaneous saturation or step deviation in sensor signals. This, in turn, triggers erroneous compensation actions in the automatic thickness control system. Although increasing network bandwidth or optimizing transmission priority can reduce communication latency, it fails to address the underlying physical process and logical command. The inherent clock mismatch between commands still makes it difficult to eliminate the random performance deviation caused by network jitter. For example, Chinese invention patent CN106597948A discloses an IoT micro embedded terminal that uses multi-channel isolated power supply, digital isolators and optocoupler isolation to improve electrical reliability and provides WiFi and RS485 networking interfaces. However, in high-stress, fast-paced rolling scenarios, the technical logic is limited to the instruction processing flow driven by the communication protocol stack. The configuration parameter loading action depends on the arrival time of network packets. It lacks a mechanism to transform physical field characteristics into execution layer logic triggers and cannot achieve synchronization between measurement and control logic and physical process at the microsecond level.

[0004] Therefore, the technical problem to be solved by this invention is how to achieve physical-level precise synchronization between the reconfiguration of IoT terminal configuration parameters and the moment of production process engagement in complex working conditions with network transmission jitter and high-intensity mechanical noise. Summary of the Invention

[0005] This invention provides a remote configuration method for embedded terminals based on the Internet of Things, comprising the following steps: Step 101: The target embedded terminal receives the configuration parameter set issued by the remote management platform and stores the configuration parameter set in the preparation buffer of the local non-volatile storage space. The configuration parameter set is in an activation state in the preparation buffer. Step 102: During the idle interval of the sensing operation, the target embedded terminal acquires the environmental background signal sequence through the high-frequency sampling unit and calculates the statistical energy distribution law of the environmental background signal sequence in the discrete time domain. Based on this, a dynamic trigger threshold is established to establish a dynamic alignment relationship between the sensing logic of the target embedded terminal and the physical environment background noise. Step 103: The target embedded terminal acquires the transient signal stream generated when the physical event is triggered in real time, and performs real-time root mean square energy feature extraction on the transient signal stream using a sliding window of preset length, so as to generate a transient signal feature vector containing transient amplitude and energy change rate features. Step 104: The target embedded terminal compares the transient signal feature vector with the dynamic trigger threshold in real time. When the amplitude of the transient signal feature vector exceeds the dynamic trigger threshold and the duration of the transient signal feature vector above the dynamic trigger threshold reaches the preset logic judgment period, the rising edge jump point of the transient signal feature vector is locked to identify the synchronous triggering time of the physical event. Step 105: In response to the synchronization trigger moment, the target embedded terminal executes the atomic loading instruction to overwrite the configuration parameter set pre-stored in the preparation buffer onto the current control logic stack in the run register, so that the perception logic of the target embedded terminal and the production rhythm of physical events can be synchronized in time.

[0006] Preferably, the step 102 of establishing the dynamic trigger threshold includes: the target embedded terminal acquiring the noise base current in the environmental background signal sequence; calculating the statistical mean and fluctuation range of the noise base current over multiple consecutive sampling periods; and performing a linear combination of the statistical mean and fluctuation range according to a preset weighting ratio stored in the target embedded terminal to calculate the dynamic trigger threshold.

[0007] Preferably, step 103, which involves using a sliding window of a preset length to perform real-time root mean square energy feature extraction on the transient signal stream, includes: the target embedded terminal mapping the transient signal stream to a discrete time domain space and establishing a time domain window that covers the complete envelope of the event pulse; performing root mean square operation within the time domain window to obtain the root mean square value of the transient energy of the transient signal stream; calculating the first derivative of the root mean square value of the transient energy with respect to the sampling time, and using the root mean square value of the transient energy and its first derivative to form a transient signal feature vector.

[0008] Preferably, the synchronous triggering moment of the physical event identified in step 104 includes: the target embedded terminal performs level transition determination on the transient signal feature vector through hysteresis comparison logic; when it is identified that the amplitude of the transient signal feature vector exceeds the dynamic triggering threshold and the duration of the transient signal feature vector above the dynamic triggering threshold exceeds the preset anti-interference duration threshold, the moment when the transient signal feature vector first exceeds the dynamic triggering threshold is confirmed as the synchronous triggering moment. While performing step 105, the target embedded terminal also performs perception accuracy compensation: the target embedded terminal retrieves the corresponding correction gradient according to the physical characteristic parameters received synchronously with the configuration parameter set, and uses the evolution operator based on cumulative deformation work to perform secondary correction on the sampling gain and filter cutoff frequency in the configuration parameter set.

[0009] Preferably, a dynamic trigger threshold The computational logic follows these rules: ,in, The energy statistical mean of the noise base current over the preset observation period. The fluctuation range of the noise base current. This is the confidence adjustment coefficient preset in the target embedded terminal.

[0010] Preferably, the execution of the atomic loading instruction in step 105 includes: the target embedded terminal starting a hardware interrupt service routine, and in a state where other peripherals are prohibited from accessing the running register, moving the configuration parameter set from the preparation buffer to the running register; after the moving is completed, clearing the sampling buffer of the target embedded terminal's sensing channel, so that subsequent sensing data are all processed based on the updated configuration parameter set.

[0011] Preferably, after executing step 105, the target embedded terminal further includes confirming the effectiveness of the execution parameters: the target embedded terminal generates a status reporting data packet containing a configuration loading identifier and a timestamp data of the synchronization trigger time, and asynchronously sends the status reporting data packet to the remote management platform through the IoT communication link.

[0012] Preferably, before executing step 101, the target embedded terminal further includes performing configuration parameter security verification: the target embedded terminal performs asymmetric encryption signature verification and legality boundary check on the configuration parameter set to determine whether the values ​​of each parameter in the configuration parameter set are within the preset hardware security threshold range. Step 102, obtaining the environmental background signal sequence, includes: the target embedded terminal uses a high-pass filtering algorithm to filter out the DC bias component in the original signal, and according to the preset device operating frequency, uses a notch filter to remove resonance interference of a specific frequency to extract the environmental background signal sequence.

[0013] An embedded terminal device based on the Internet of Things (IoT) includes: a memory for storing computer program instructions; and a processor for executing the computer program instructions to implement the steps of the IoT-based embedded terminal remote configuration method.

[0014] An Internet of Things (IoT) based embedded terminal medium stores a computer program, which, when executed by a processor, implements the steps of the IoT-based embedded terminal remote configuration method.

[0015] Compared with existing technologies, the present invention, based on the Internet of Things (IoT) embedded terminal remote configuration method, device, and medium, has the following advantages: 1. In embedded terminal remote configuration, by decoupling the acquisition and loading of remote configuration parameters in the time dimension, the stress step signal generated by the metal part entering the roll is used as the physical clock source for configuration reconstruction. This transforms the instruction activation logic from being driven by the communication protocol stack to being directly triggered by the physical field of the production site, eliminating the constraint of wireless network transmission jitter on the determinism of parameter switching, and achieving timing alignment between the sensing characteristics of the measurement and control system and the production process rhythm at the microsecond level.

[0016] 2. The background energy envelope extracted during the unloaded rotation of the rolls is used to dynamically correct the trigger threshold. Combined with the multi-level correction gradient preset according to the cold hardening effect of the head of the metal part to be rolled and the evolution operator based on the cumulative deformation work, the signal conditioning gain is synchronously shifted with the thermal evolution and working condition deviation during the rolling process. This eliminates the degradation of sensing accuracy caused by the fixed trigger threshold or the mismatch between the preset parameters and the real-time physical field, and ensures that the data acquisition link maintains a high degree of linearity and response consistency at the moment of bite impact and in the subsequent steady-state rolling stage.

[0017] 3. By reusing the no-load vibration signal of the rolling mill stand to invert the health fingerprint of the sensor link, and in conjunction with the secondary arbitration logic for stress attenuation echo characteristics, a dual interlocking protection against environmental interference and hardware damage is built before parameter loading. This ensures that configuration reconfiguration is activated only when the sensing path is intact and a real bite event is identified, thereby improving the self-diagnostic capability of IoT nodes in complex electromechanical environments and ensuring the data security of the input source of the production control system. Attached Figure Description

[0018] Figure 1 This is a flowchart of the physical field-triggered remote configuration method for embedded terminals according to the present invention; Figure 2 This is a system principle block diagram of the collaboration between the digital instruction domain and the physical field domain of the present invention. Detailed Implementation

[0019] 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.

[0020] It should be noted that all directional and positional terms used in this invention, such as: up, down, left, right, front, back, vertical, horizontal, inner, outer, top, bottom, transverse, longitudinal, center, etc., are only used to explain the relative positional relationship and connection between components in a specific state (as shown in the accompanying drawings). They are only for the convenience of describing this invention and do not require that this invention be constructed and operated in a specific orientation. Therefore, they should not be construed as limiting this invention. In addition, the descriptions of "first," "second," etc., in this invention are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated.

[0021] In the description of this invention, unless otherwise explicitly specified and limited, the terms installation, connection, and linking should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to mechanical connections; they can refer to direct connections or indirect connections through an intermediate medium; they can refer to the internal communication between two components. For those skilled in the art, the specific meaning of the above terms in this invention can be understood according to the specific circumstances.

[0022] In the description of this specification, references to the terms "an embodiment," "some embodiments," "illustrative embodiments," "examples," "specific examples," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example, and the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0023] A method, device, and medium for remote configuration of an embedded terminal based on the Internet of Things (IoT) are disclosed. In the context of continuous metal rolling production, the configuration parameters of the IoT embedded terminal need to be switched at the millisecond level according to the physical characteristics of the metal workpiece to be rolled. Due to link jitter during data transmission in industrial wireless networks or time-sensitive networks, the arrival time of the configuration command at the terminal is uncertain. If the configuration parameters take effect during the metal biting process, it will cause an instantaneous step deviation in the sampled data, thereby triggering the erroneous compensation action of the automatic thickness control system. To solve this problem, this invention provides a method to decouple the acquisition of configuration parameters from the loading action in the time dimension, using the stress characteristics generated by the physical field of the production site as the physical clock source for parameter switching; Target embedded The terminal receives the configuration parameter set issued by the remote management platform, stores it in the preparatory buffer of the local non-volatile storage space, and is in an inactive state in the preparatory buffer. The target configuration parameter set includes the configuration base value, multi-level correction gradient, signal conditioning gain value, and filter cutoff frequency. The IoT gateway extracts the carbon content percentage value of the current metal part to be rolled by accessing the steel coil production list of the production management system, and encapsulates it as a physical characteristic parameter in the data packet header field of the configuration parameter set. It is then synchronously distributed to the target embedded terminal along with the parameter set. When the IoT gateway detects the empty gap after the previous coil of metal has been rolled, it pre-transmits the target configuration parameter set for the next coil of metal to the cache unit of the target embedded terminal.

[0024] Because of the mechanical background noise caused by bearing wear or rotational imbalance during the no-load rotation of the rolling mill stand, a fixed trigger threshold is prone to false triggering or response lag. During the no-load interval of the sensing operation, the target embedded terminal acquires the environmental background signal sequence through a high-frequency sampling unit. The sampling frequency of this high-frequency sampling unit is set in the range of 10kHz to 50kHz. The target embedded terminal calculates the statistical energy distribution law of the environmental background signal sequence in the discrete time domain, and establishes the dynamic trigger threshold accordingly. Specifically, the target embedded terminal acquires the noise fundamental current in the environmental background signal sequence and calculates the statistical mean of the energy of the noise fundamental current over multiple consecutive sampling periods. and fluctuation range Adjust according to the preset confidence level coefficient Calculate the dynamic trigger threshold Dynamic trigger threshold The calculation formula is as follows: ,in, For dynamic trigger thresholds; σ represents the statistical mean energy of the noise base current within a preset observation period; σ represents the fluctuation range of the noise base current; α represents the confidence adjustment coefficient preset in the target embedded terminal, with a value ranging from 1.2 to 1.8. This calculation process establishes the trigger threshold above the peak of the energy envelope of the current mechanical state, achieving dynamic alignment between the sensing logic and the background noise of the physical environment. The confidence adjustment coefficient α is selected according to the signal-to-noise ratio (SNR) of the environmental background signal sequence. The target embedded terminal statistically analyzes the energy distribution of the environmental background signal sequence during the idle interval, calculates the average energy and peak energy of the noise base current, and sets the confidence adjustment coefficient α to a calibration value that is negatively correlated with SNR. When the SNR is higher than 25 dB, α is 1.2 to 1.4; when the SNR is lower than 15 dB, α is 1.6 to 1.8. By establishing a quantitative correlation between α and the on-site mechanical noise intensity, the dynamic trigger threshold is set. Positioned above the background noise envelope and close to the starting point of the physical event trigger signal, it improves the system's sensitivity to metal bite events.

[0025] After entering the rolling stage, the target embedded terminal acquires the transient signal stream generated by the metal biting in real time, and the target embedded terminal uses a preset length of to A sliding window performs real-time root-mean-square energy feature extraction on the transient signal stream, generating a transient signal feature vector containing transient amplitude and energy change rate features. The target embedded terminal maps the transient signal stream to a discrete time domain space, establishing a time domain window covering the complete envelope of the event pulse. Within this time domain window, root-mean-square operations are performed to obtain the transient energy root-mean-square value. Simultaneously, the first derivative of this transient energy root-mean-square value with respect to the sampling time is calculated. This transient energy root-mean-square value and its first derivative together form the transient signal feature vector, used to characterize the stress wave characteristics of the metal biting state. The target embedded terminal then combines the transient signal feature vector with a dynamic trigger threshold. Real-time comparison is performed when the amplitude of the transient signal feature vector exceeds the dynamic trigger threshold. Furthermore, the transient signal feature vector is at the dynamic trigger threshold. When the duration of the above reaches the preset logic determination period, the target embedded terminal locks the rising edge of the transient signal feature vector to identify the synchronous triggering moment of the physical event. This logic determination period serves as an anti-interference duration threshold to filter high-frequency random impact noise.

[0026] The procedure for determining the synchronous trigger time includes algorithm delay pre-compensation processing, and the target embedded terminal determines the trigger time based on the sliding window length. With high frequency sampling unit sampling frequency Determine the processing delay of the algorithm , Values The processor detected that the transient signal feature vector exceeded the dynamic trigger threshold. At that time, the system timestamp is recorded using a hardware capture unit, and the algorithm processing delay is subtracted. To restore the true moment of the physical event, the atomized loading instruction completes parameter overwriting during the rising phase before the physical stress wave reaches the preset energy ratio, ensuring that the switching of sensing parameters is consistent with the timing of the physical deformation process. In response to this synchronization trigger moment, the target embedded terminal starts the hardware interrupt service routine and executes the atomized loading instruction. During the execution of this instruction, the processor prohibits other peripherals from accessing the running register, moves the configuration parameter set pre-stored in the preparation buffer to the current control logic stack of the running register, and after loading is completed, clears the sampling buffer of the sensing channel so that subsequent sensing data is processed based on the updated configuration parameter set. This process realizes the synchronous switching of the sensing logic of the target embedded terminal with the physical production rhythm.

[0027] The execution of atomic loading instructions employs a hardware register mapping switching method. The target embedded terminal internally is divided into a running logic area and a preparatory storage area. Upon response to the synchronization trigger, the hardware interrupt service routine modifies the processor status register, redirecting the control logic addressing base address to the preparatory buffer storage space. This achieves configuration parameter replacement within a single instruction cycle. This process involves only five to eight machine cycles of pointer switching operations, does not occupy the sensing channel's sampling bus, and employs a double-buffered circular queue mechanism. During pointer switching, sampling points are continuously written to the spare buffer, ensuring that sensing data is not interrupted or lost during configuration reconstruction. The processor, through bus arbitration logic, prohibits direct memory access to the channel from other peripherals. The line occupies the data path of the running register to avoid data conflicts or floating control commands during parameter loading, ensuring that the real-time control logic of the terminal during parameter reconstruction phase is not interrupted. Due to the cold hardening effect at the head of the metal part to be rolled, its deformation resistance exhibits a transient peak in the initial bite stage. To prevent saturation distortion of the sampling signal, the target embedded terminal extracts the energy peak of the instantaneous load signal within a preset calibration period after parameter reconstruction. The target embedded terminal compares the deviation of this energy peak from the configured reference value and matches the target correction value from a multi-level correction gradient accordingly, performing online fine-tuning of the sampling gain in the running register. In addition, during the steady-state rolling stage, the target embedded terminal calculates the cumulative deformation work of the metal part to be rolled in real time. The target embedded terminal performs time integration calculation on the instantaneous rolling force signal and calculates the cumulative deformation work. An evolution correction vector is determined with a preset compensation operator. This evolution correction vector is then used to perform step-by-step updates on the signal processing parameters in the operation register to adapt to the physical response drift caused by the thermal expansion of the rolls.

[0028] To address the risk of false triggering under extreme operating conditions, the target embedded terminal introduces a secondary arbitration mechanism. Within a preset delay window after the physical trigger pulse is generated, the target embedded terminal extracts the attenuated wave fingerprint from the instantaneous load signal, which reflects the intrinsic vibration characteristics after the metal comes into contact with the roll. If the attenuated wave fingerprint matches a pre-stored contact feature fingerprint, the loading action of the configuration parameter set is executed. If the fingerprint does not match, it is determined to be mechanical vibration interference, and the configuration loading path is blocked. At the same time, the target embedded terminal reuses the rotation-associated signal generated during the roll's unloaded rotation phase to analyze its spectral evolution characteristics relative to the pre-stored reference signal. If the analysis results show that the signal link health is lower than a preset threshold, the link anomaly is reported to the remote management platform. Changyuan Information ensures the operational security of the measurement and control system. After executing the atomic loading instruction, the target embedded terminal generates a status reporting data packet, including a configuration loading success identifier and a timestamp of the synchronization trigger time. The target embedded terminal asynchronously sends this status reporting data packet to the remote management platform via the IoT communication link to complete the configuration activation confirmation. During the security verification process, the target embedded terminal initiates a hardware-level data legality review procedure to establish the physical boundary constraints of the configuration parameter set. The system uses a built-in security verification algorithm to perform integrity verification on the digital digest carried by the configuration parameter set, and uses a logical comparison operator to compare the received signal conditioning gain value with the gain safety upper limit preset in the non-volatile memory. Perform numerical comparisons and determine whether the filter cutoff frequency is within the frequency range allowed by hardware clock constraints. In this implementation scenario, the gain safety upper limit is... The setting is 60dB. If any configuration parameter is detected to exceed its corresponding hardware safety threshold range, the internal configuration interlocking logic is triggered and an alarm status word is generated and sent to the remote management platform to prevent the perception logic from failing due to illegal parameter writing.

[0029] Example 1: In a high-speed continuous rolling production line with an annual output of 2,000,000 tons, the measurement and control system faces communication link jitter caused by industrial wireless networks or time-sensitive networks. This jitter makes the arrival time of configuration commands issued by the remote management platform at the target embedded terminal uncertain. If the command takes effect at a time that deviates from the physical time when the metal enters the rolls, it will cause a step distortion in the collected metal thickness data, which in turn will cause the automatic thickness control system to make incorrect adjustments. To resolve this conflict, the target embedded terminal executes a silent preprocessing procedure, using the background noise generated by the mill stand during the unloaded rotation phase as a calibration source. The environmental background signal sequence is acquired through a high-frequency sampling unit, and the energy distribution characteristics of the sequence within the time-domain window are calculated to determine the dynamic trigger threshold. Dynamic trigger threshold The calculation formula is as follows: ,in, The dynamic trigger threshold is expressed in volts. The energy of the noise base current over a preset observation period is the statistical mean, expressed in volts. The fluctuation range of the noise base current is expressed in volts. The confidence adjustment coefficient is preset in the target embedded terminal, and is set to 1.5 in this implementation scenario. This feature extraction process transforms randomly distributed mechanical noise into a defined energy boundary, providing a physical alignment benchmark for subsequent configuration parameter presentation.

[0030] After entering the rolling phase, the target embedded terminal acquires the transient signal stream generated at the moment the metal bites in real time, and extracts the transient signal feature vector composed of transient amplitude and energy change rate. When its amplitude crosses the dynamic trigger threshold... Furthermore, when maintaining an anti-interference duration exceeding 10ms in the time domain, the target embedded terminal locks the level transition point as the synchronization trigger moment, initiates a hardware interrupt to execute the atomic loading instruction, and directly overwrites the target configuration parameter set pre-stored in the cache unit into the running register. This transfers the driving force for configuration activation from the uncertain communication protocol stack to the physical stress field of the production site, eliminating the interference of network latency on the timing characteristics of the sensing system, and enabling the configuration switching action and the metal deformation process to achieve synchronous operation at the microsecond level. For the initial load impact caused by the cold hardening effect at the metal part's head, the target embedded terminal identifies the energy peak within a preset period after configuration activation and compares it with the configuration reference value. It then selects the corresponding target correction value from the preset multi-level correction gradient and performs real-time compensation on the signal gain coefficient in the running register. After entering steady-state rolling, the system uses a time integration operator to calculate the cumulative deformation work of the metal part to be rolled. According to the cumulative deformation work The evolution trend of the magnitude is observed, and the filter cutoff frequency and gain parameters in the running register are finely adjusted in steps. This process solves the sensing mismatch caused by the thermal expansion of the roll and the cumulative effect of the physical process through the dynamic evolution of the parameters. A remote configuration mode with physical stress characteristics as clock pulses is established. By real-time calibration of the background environment energy and atomized loading of the physical trigger vector, a closed-loop response chain with physical determinism is constructed in a nondeterministic communication environment, so as to realize the stable migration of the sensing characteristics of the measurement and control system with the evolution of the production conditions.

[0031] Example 2: In a simulated test environment of a high-speed continuous rolling production line with an annual output of 2,000,000 tons, a closed-loop verification platform was constructed using a load simulator that includes an IoT gateway, a target embedded terminal, and a hydraulic actuator capable of outputting sudden physical stress wave signals. This verified the engineering effectiveness of the remote configuration method in eliminating the impact of communication link jitter on sensing accuracy. The target embedded terminal uses a high-frequency sampling unit with 16-bit analog-to-digital conversion accuracy and a sampling frequency set to 50kHz for data acquisition. The data source is the vibration load sequence generated in real time by the physical experimental platform. To simulate the interference of the real industrial electromagnetic environment on the signal link, Gaussian white noise with a signal-to-noise ratio of 20dB and power frequency harmonics with a frequency of 50Hz are actively superimposed on the output signal of the load simulator.

[0032] Regarding the setting of the confidence adjustment coefficient α for the key parameter in the experiment, the technical trade-off lies in achieving a balance between the system's false trigger rate and real-time response. If the value of α is too low, the random energy fluctuations of the mechanical background noise can easily exceed the threshold, causing premature parameter loading; if... Higher values ​​trigger dynamic thresholds Approaching the peak signal energy will cause physical lag in the decision logic. By performing gradient calibration experiments within the background noise fluctuation range, it was determined that when α is in the range of 1.2 to 1.8, the system can isolate more than 98.5% of random mechanical shocks and keep the response delay below 1ms. A typical value of 1.5 was selected as the calibration parameter for this experiment. At the same time, the network latency of the configuration command issued by the IoT gateway was set to fluctuate randomly between 50ms and 200ms to simulate the objective challenges of nondeterministic transmission links.

[0033] In the experiment, the traditional timed triggering method based on network time protocol was set as control group 1, the method using fixed energy threshold judgment was set as control group 2, and the method using the dynamic triggering threshold and physical event driven loading of the present invention was set as the experimental group of the present invention. The alignment accuracy of configuration switching was characterized by the deviation of the sensing data at the moment of metal bite. Table 1 is a comparison table of configuration synchronization performance test data. The table records the average alignment deviation value and signal step error of the three schemes under different network jitter intensities. The alignment deviation value is determined by the difference between the synchronization triggering time and the time when the load simulator outputs the step signal.

[0034] Table 1: Comparison of Synchronization Performance Test Data Analysis of the data in Table 1 shows that as network latency jitter increased from 52.4 ms to 300.0 ms, the alignment deviation of control group 1 exhibited a quasi-linear growth trend with jitter intensity, indicating that the timing triggering mechanism was completely controlled by the uncertainty of the communication link. In contrast, control group 2, constrained by a fixed threshold to accommodate background noise, maintained an alignment deviation above 12.5 ms despite being unaffected by the network. The experimental group of this invention, through dynamic triggering thresholds… Real-time comparison with the transient signal feature vector consistently maintained a alignment deviation within 1.05 ms, and the signal step error was suppressed to below the system background level of 6.0 mV. To verify the rationality of the confidence adjustment coefficient α range and identify performance inflection points, the system response characteristics were tested when α exceeded the preset range. When α decreased to 0.5, the background signal energy frequently crossed the calculated range. The value caused the system to trigger an erroneous action during the no-load phase, resulting in the configuration parameters being loaded before the metal was even engaged; and when After being increased to version 3.0, due to the dynamic trigger threshold... The rapid growth zone of the bite stress wave was exceeded, and the system waited for the signal to reach its peak before triggering, causing the alignment deviation to suddenly increase from 0.82ms to 35.4ms. This experiment, by comparing the response accuracy of different configuration loading modes under a gradient network jitter environment, confirmed that the proposed solution utilizes the noise base current. With fluctuation range The technical approach of real-time correction of trigger thresholds can transform uncertain communication layer latency into deterministic execution layer physical response, solving the perception mismatch problem caused by the deviation of parameter activation point from the physical production rhythm under high-speed rolling conditions.

[0035] Example 3: This example combines Figures 1 to 2 This document describes a remote configuration method, device, and medium for an embedded terminal based on the Internet of Things (IoT). Figure 1As shown, in step 101, the target embedded terminal receives the configuration parameter set issued by the remote management platform and stores it in the preparation buffer of the local non-volatile storage space, so that the configuration parameter set is in an activated state in the preparation buffer. Then, in step 102, during the idle interval of the sensing operation, the target embedded terminal obtains the environmental background signal sequence through the high-frequency sampling unit and calculates its statistical energy distribution law in the discrete time domain to establish the dynamic trigger threshold, thereby establishing the dynamic alignment relationship between the sensing logic and the physical environment background noise. Then, in step 103, the transient signal stream generated when the physical event is triggered is obtained in real time, and a sliding window of preset length is used to analyze the transient signal. The signal stream performs real-time root mean square energy feature extraction to generate a transient signal feature vector containing transient amplitude and energy change rate features. Based on this, step 104 compares the transient signal feature vector with the dynamic trigger threshold in real time. When the amplitude intensity exceeds the threshold and the duration reaches the logic judgment period, the rising edge jump point of the transient signal feature vector is locked to identify the synchronous trigger moment of the physical event. Finally, step 105 is executed in response to the synchronous trigger moment to execute the atomic loading instruction, which overwrites the pre-stored configuration parameter set onto the current control logic stack in the running register, so that the production rhythm of the sensing logic and the physical event can achieve time synchronization switching at the microsecond level.

[0036] like Figure 2 As shown, on the configuration source side in the digital command domain, the remote management platform executes a distribution operation, transmitting the command to the IoT gateway. The configuration package generated by the IoT gateway is transmitted to the target embedded terminal and stored in the terminal's internal preparatory buffer in a standby state. Simultaneously, on the trigger source side in the physical field domain, the mechanical impact generated by the mill stand acts on the piezoelectric sensor, which outputs a trigger signal and transmits it directly to the high-frequency sensing logic module of the target embedded terminal. After receiving the trigger signal from the physical side, the target embedded terminal, through internal logic processing, drives the configuration data in the preparatory buffer to take effect, thereby realizing remote configuration closed-loop control based on physical field triggering.

[0037] Example 4: In the measurement and control scenario of a continuous rolling production line for carbon structural steel and low-alloy high-strength steel, the target embedded terminal needs to set initial signal processing values ​​for metal parts with different yield strengths. Since the metal deformation resistance increases with increasing carbon content, if the multi-level correction gradient is qualitatively set based solely on physical perception, it is easy to cause sampling overflow in high-yield-strength steel at the moment of biting. To establish a gain quantization benchmark in the configuration parameter set, the target embedded terminal executes a parameter calibration procedure. In the offline stage before the system is put into production, the signal conditioning gain value is established based on the intrinsic hardness coefficient of the metal part to be rolled. The calculation formula for the signal conditioning gain value is as follows: ,in, This is the signal conditioning gain value, expressed in decibels (dB). This is the reference gain constant, which is set to 40dB in this implementation scenario; This is the hardness influence factor, with a value of 15dB / ; The carbon content percentage of the metal part to be rolled is used. The calibration logic pre-generates a multi-level correction gradient for metal parts with a carbon content between 0.1wt% and 0.8wt%. When the carbon content increases by 0.1wt%, the signal conditioning gain value is reduced by 1.5dB so that the dynamic range of the front-end amplifier circuit covers the starting peak of the stress wave.

[0038] During the steady-state rolling stage after metal infeed, to correct the perception deviation of the reduction force caused by the thermal expansion of the rolls, the system initiates a dynamic evolution mapping process, which obtains the cumulative deformation work by integrating the instantaneous rolling force signal over time. And based on the cumulative deformation work The evolution correction vector is calculated using the following formula: ,in, This is the evolution correction vector, used for step-by-step adjustment of the offset compensation term in the run register, in millimeters; The evolution compensation coefficient is determined by the thermal expansion coefficient of the roll material, and is set to 0.002 mm / s in this implementation scenario. ; The cumulative deformation work is measured in megajoules. This mapping process maintains a linear proportional relationship between the updated configuration parameters and the cumulative energy input of the physical process, eliminating temperature drift interference at the sensing zero point. In addition, the target embedded terminal determines a preset safety boundary through quantization calibration, which is defined as the signal sampling value accounting for 85% of the full scale of the analog-to-digital converter. When the energy peak value in the calibration cycle exceeds this value, the gain degradation instruction in the multi-level correction gradient is triggered. By executing the parameter calibration and mapping procedure determined above, the target embedded terminal converts the original steel grade information into precise low-level register control quantities, so that the linearity deviation of the sensing data is controlled below 0.5% when the measurement and control system faces the production conditions of material switching.

[0039] Example 5: In a field deployment scenario for a newly built metal rolling production line, the target embedded terminal completes a pre-deployment calibration process to build a baseline fingerprint library for the sensor link. The system conducts segmented rotational tests within the range of 50% to 100% of the rated speed on the mill stand when there is no material and the speed is not increased. The high-frequency sampling unit continuously acquires the rotational accompanying signals under different speed gradients, and applies the fast Fourier transform algorithm to calculate the energy amplitude of each frequency component, thereby establishing an initial reference vector containing the fundamental resonant frequency and the background energy envelope. ;in, , As the initial reference vector, its components For the first Normalized energy amplitude at each characteristic frequency point Given the total number of characteristic frequency points, this calibration process writes the physical intrinsic characteristics of a specific mechanical structure into a non-volatile memory, which serves as a benchmark for subsequent link health mapping.

[0040] When the system faces changes in the physical environment due to sensor replacement or major mechanical overhaul, the target embedded terminal initiates an adaptive baseline reconstruction process. This involves controlling the rolls to rotate unloaded at a constant low speed, simultaneously acquiring five cycles of instantaneous load signals as reconstruction samples. The system extracts periodic fluctuation characteristics caused by bearing clearance or roll eccentricity and updates them to the fingerprint comparison table in the cache unit, completing a controlled pulse excitation response test and recording the attenuation time constant of the stress wave in the frame's metal medium. This is used to correct the attenuated wave fingerprint determination model in the secondary arbitration logic; among which, , The decay time constant is expressed in milliseconds. for The signal energy value at time t, in joules. for The signal energy value at any given time, measured in joules, is used in this calibration procedure to align the sensing logic with the electromechanical characteristics of the physical location, enabling the system to enter a calibrated operating state.

[0041] Example 6: In a real-time configuration scenario of a cold rolling production line containing impulse noise and electromagnetic harmonic interference, the target embedded terminal executes a signal stability calibration procedure to establish an anti-interference duration threshold, and the system obtains the current sampling frequency of the high-frequency sampling unit. The filtering depth of the sampling point is determined based on the shortest impact pulse width generated by the mechanical transmission chain at its highest speed. The formula for calculating the anti-interference duration threshold is: ,in, This is the threshold for anti-interference duration, in milliseconds. The threshold for counting consecutive out-of-limit sampling points is set to 8 in this implementation scenario. The sampling frequency is expressed in kilohertz. This calibration procedure filters out microsecond-level transient interference to prevent the atomized loading command's operational state from jumping due to a single sampling noise point. When the system faces production conditions where the rolling speed dynamically fluctuates between 5 m / s and 20 m / s, the target embedded terminal executes a timeliness guarantee procedure for parameter evolution, determining the calibration cycle of configuration parameters by sensing the envelope stability of the instantaneous load signal. Calculate the calibration cycle The formula is ,in, The calibration period is in seconds. The specified length of the metal part to be rolled is 1.0m in this implementation scenario. For real-time rolling speed, measured in meters per second, this procedure enables the system to fine-tune the signal conditioning gain at constant spatial intervals as the metal head deformation approaches a steady state, eliminating the cumulative deformation work caused by speed fluctuations. The integral deviation enables the parameter evolution of the target embedded terminal to be adaptively corrected as the metal plastic deformation process progresses.

[0042] In specific application scenarios aimed at improving the accuracy of physical event recognition, the target embedded terminal executes a fingerprint consistency determination procedure based on a cross-correlation function to verify the validity of the physical trigger pulse. After extracting the attenuated wave fingerprint from the instantaneous load signal, the system converts it into a discrete sequence. And retrieve the pre-stored reference contact fingerprint sequence from local memory. Calculate discrete sequence fingerprint sequence in contact with the reference Maximum correlation coefficient within the cross-correlation window The calculation formula is as follows: ,in, The maximum correlation coefficient; The extracted attenuated wave discrete sequence; The pre-stored baseline contact fingerprint discrete sequence; For time-shift parameters; when the maximum correlation coefficient When the similarity exceeds the preset threshold of 0.85, it is determined that the physical trigger pulse is generated by the metal bite event, and the execution path of the atomic loading instruction is released; otherwise, it is determined to be an interference signal and the update action of the running register is blocked.

[0043] The embodiments of this application have been described above with reference to the accompanying drawings. Unless otherwise specified, the embodiments and features in the embodiments of this application can be combined with each other. This application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit of this application and the scope of protection of this invention, and all of these forms are within the protection scope of this application.

Claims

1. A remote configuration method for embedded terminals based on the Internet of Things, characterized in that, Includes the following steps: Step 101: The target embedded terminal receives the configuration parameter set issued by the remote management platform and stores the configuration parameter set in the preparation buffer of the local non-volatile storage space. The configuration parameter set is in an activation state in the preparation buffer. Step 102: During the idle interval of the sensing operation, the target embedded terminal acquires the environmental background signal sequence through the high-frequency sampling unit and calculates the statistical energy distribution law of the environmental background signal sequence in the discrete time domain. Based on this, a dynamic trigger threshold is established to establish a dynamic alignment relationship between the sensing logic of the target embedded terminal and the physical environment background noise. Step 103: The target embedded terminal acquires the transient signal stream generated when the physical event is triggered in real time, and performs real-time root mean square energy feature extraction on the transient signal stream using a sliding window of preset length, so as to generate a transient signal feature vector containing transient amplitude and energy change rate features. Step 104: The target embedded terminal compares the transient signal feature vector with the dynamic trigger threshold in real time. When the amplitude of the transient signal feature vector exceeds the dynamic trigger threshold and the duration of the transient signal feature vector above the dynamic trigger threshold reaches the preset logic judgment period, the rising edge jump point of the transient signal feature vector is locked to identify the synchronous triggering time of the physical event. Step 105: In response to the synchronization trigger moment, the target embedded terminal executes the atomic loading instruction to overwrite the configuration parameter set pre-stored in the preparation buffer onto the current control logic stack in the run register, so that the perception logic of the target embedded terminal and the production rhythm of physical events can be synchronized in time.

2. The method for remote configuration of an embedded terminal based on the Internet of Things according to claim 1, characterized in that, Step 102, establishing the dynamic trigger threshold, includes: the target embedded terminal acquiring the noise base current in the environmental background signal sequence; calculating the statistical mean and fluctuation range of the noise base current over multiple consecutive sampling periods; and performing a linear combination of the statistical mean and fluctuation range according to a preset weighting ratio stored in the target embedded terminal to calculate the dynamic trigger threshold.

3. The method for remote configuration of an embedded terminal based on the Internet of Things according to claim 1, characterized in that, Step 103 involves performing real-time root mean square energy feature extraction on the transient signal stream using a sliding window of a preset length. This includes: the target embedded terminal mapping the transient signal stream to a discrete time domain space and establishing a time domain window that covers the complete envelope of the event pulse; performing root mean square operation within the time domain window to obtain the root mean square value of the transient energy of the transient signal stream; calculating the first derivative of the root mean square value of the transient energy with respect to the sampling time, and using the root mean square value of the transient energy and its first derivative to form a transient signal feature vector.

4. The method for remote configuration of an embedded terminal based on the Internet of Things according to claim 1, characterized in that, Step 104 identifies the synchronous trigger moment of the physical event, including: the target embedded terminal uses hysteresis comparison logic to determine the level transition of the transient signal feature vector; when the amplitude of the transient signal feature vector exceeds the dynamic trigger threshold and the duration of the transient signal feature vector above the dynamic trigger threshold exceeds the preset anti-interference duration threshold, the moment when the transient signal feature vector first exceeds the dynamic trigger threshold is identified as the synchronous trigger moment. While executing step 105, the target embedded terminal also performs perception accuracy compensation: the target embedded terminal retrieves the corresponding correction gradient according to the physical characteristic parameters received synchronously with the configuration parameter set, and uses the evolution operator based on cumulative deformation work to perform secondary correction on the sampling gain and filter cutoff frequency in the configuration parameter set.

5. The method for remote configuration of an embedded terminal based on the Internet of Things according to claim 2, characterized in that, Dynamic trigger threshold The computational logic follows these rules: ,in, The energy statistical mean of the noise base current over the preset observation period. The fluctuation range of the noise base current. This is the confidence adjustment coefficient preset in the target embedded terminal.

6. The method for remote configuration of an embedded terminal based on the Internet of Things according to claim 1, characterized in that, Step 105, executing the atomic loading instruction, includes: the target embedded terminal starting a hardware interrupt service routine, and, with other peripherals prohibited from accessing the run register, moving the configuration parameter set from the preparation buffer to the run register; after the movement is completed, clearing the sampling buffer of the target embedded terminal's sensing channel, so that subsequent sensing data are all processed based on the updated configuration parameter set.

7. The method for remote configuration of an embedded terminal based on the Internet of Things according to claim 1, characterized in that, After executing step 105, the target embedded terminal also includes confirming the effectiveness of the execution parameters: the target embedded terminal generates a status reporting data packet containing a configuration loading success identifier and a timestamp data of the synchronization trigger time, and asynchronously sends the status reporting data packet to the remote management platform through the IoT communication link.

8. The method for remote configuration of an embedded terminal based on the Internet of Things according to claim 1, characterized in that, Before executing step 101, the target embedded terminal also performs configuration parameter security verification: the target embedded terminal performs asymmetric encryption signature verification and legality boundary check on the configuration parameter set to determine whether the value of each parameter in the configuration parameter set is within the preset hardware security threshold range. Step 102 to obtain the environmental background signal sequence includes: the target embedded terminal uses a high-pass filtering algorithm to filter out the DC bias component in the original signal, and according to the preset device operating frequency, it uses a notch filter to remove resonance interference of a specific frequency to extract the environmental background signal sequence.

9. An embedded terminal device based on the Internet of Things, characterized in that, include: Memory is used to store computer program instructions; A processor for executing computer program instructions to implement the steps of the method as described in claim 1.

10. An embedded terminal medium based on the Internet of Things, characterized in that, The embedded terminal medium based on the Internet of Things stores a computer program, which, when executed by a processor, implements the steps of the method described in claim 1.

Images