Intelligent vegetable cleaning machine monitoring system and method based on Internet of Things

Abstract

The invention relates to the technical field of Internet of Things equipment monitoring, and discloses an intelligent vegetable cleaning machine monitoring system and method based on Internet of Things. The equipment state monitoring module of the system acquires equipment state information of the vegetable cleaning machine in real time, so that a user can master the basic condition of the equipment at any time. The process parameter detection module monitors water temperature, water level and turbidity parameters in the cleaning process in real time, and proper conditions are provided for vegetable cleaning. The industrial control module executes the functions of starting and stopping operation, sequential starting and stopping control, equipment interlocking control, emergency stopping control and the like, so that the operation is more intelligent and convenient. The data integration transmission module receives real-time data of the two modules and sends the data to the intelligent analysis module. And the intelligent analysis module establishes a vegetable cleaning process dynamic model by using the data and generates a cleaning parameter optimization instruction. And the cloud service module stores historical data and real-time data and also provides remote access and monitoring functions, thereby facilitating data management and remote operation.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of Internet of Things device monitoring, in particular to an intelligent vegetable cleaning machine monitoring system and method based on Internet of Things. BACKGROUND

[0002] In the field of food processing, vegetables as an important raw material, the degree of cleaning directly related to the final food quality and safety. Whether it is to make canned vegetables, frozen vegetables, or processed into vegetable juice, vegetable dry products, cleaning is an indispensable first step. If not thoroughly cleaned, residual mud, impurities and pesticides will affect the taste, quality of food, and may even cause food safety problems.

[0003] The catering industry is also highly dependent on the quality of vegetable cleaning. Restaurants, canteens and other places need a large amount of fresh, clean vegetables every day to make various dishes. Clean vegetables not only can improve the appearance and taste of dishes, but also can protect the health of consumers and maintain the reputation of catering enterprises.

[0004] Traditional vegetable cleaning methods have many problems. Manual cleaning is inefficient and difficult to meet the needs of large-scale food processing and catering services, and labor costs continue to rise, increasing the operating burden of enterprises. At the same time, the quality of manual cleaning is difficult to guarantee consistency and may not be thoroughly cleaned. Some existing vegetable cleaning equipment, although it improves the cleaning efficiency to some extent, but there are serious deficiencies in equipment monitoring and intelligent control.

[0005] Traditional vegetable cleaning machines have limited ability in equipment state monitoring. It is difficult for operators to accurately understand whether the equipment is in a normal power-on state in real time, and it is also difficult to know whether the equipment has failed in time. For example, when a key component inside the device is worn or damaged, it may not immediately show obvious abnormalities, but continued operation will exacerbate the equipment failure, even stop, affecting production progress. Moreover, the traditional cleaning machine lacks support for remote control mode, and in some large food processing enterprises or chain catering institutions, it is impossible to realize centralized management and control of cleaning machines distributed in different areas, reducing management efficiency.

[0006] In terms of cleaning process parameter control, the traditional cleaning machine relies entirely on the experience of the operator to set parameters such as water temperature, water level and turbidity. The experience levels of different operators vary, resulting in unstable cleaning effect. For example, too high water temperature may damage the nutritional content and taste of vegetables, too low water level may not be able to fully clean the vegetables, and inaccurate turbidity monitoring may not be able to replace the cleaning water in time, thus affecting the cleaning quality.

[0007] In terms of operation control, the start-stop operation, sequential start-stop control, and device interlock control of traditional cleaning machines are relatively simple, lack flexibility and intelligence. In the event of an emergency, the response speed of the emergency stop control may be slow, and accidents may not be effectively avoided.

[0008] In terms of data management, traditional cleaning machines lack the ability to integrate and transmit data and intelligent analysis. A large amount of data generated during the cleaning process, such as device running time, cleaning frequency, fault records, etc., cannot be effectively collected and analyzed. This makes it difficult for enterprises to extract valuable information from data, unable to optimize the cleaning process, and not conducive to equipment maintenance and management.

[0009] The Internet of Things technology has developed rapidly and gradually penetrated into various industries. The Internet of Things connects various devices, objects, and the Internet to achieve data transmission and exchange, thereby realizing intelligent management and control. In the industrial field, the Internet of Things technology enables devices in the factory to upload real-time operation data, allowing management personnel to remotely monitor device status, detect potential faults in advance and perform maintenance, greatly improving production efficiency and device reliability. In the smart home field, the Internet of Things enables home devices to interconnect, allowing users to remotely control home appliances, lighting, and other devices through mobile terminals, improving the convenience and comfort of life.

[0010] The Internet of Things technology also plays an important role in agriculture, medicine, transportation, and other industries. In agriculture, sensors monitor soil moisture, temperature, nutrients, and other information to achieve precise irrigation and fertilization; in the medical field, wearable devices transmit patient health data to doctors in real time through the Internet of Things, facilitating remote diagnosis and treatment; in the transportation field, the Internet of Things technology is used for intelligent traffic management to optimize traffic flow and intelligently schedule vehicles.

[0011] The successful application of the Internet of Things technology in various industries provides new ideas and methods for solving the pain points of vegetable cleaning machine monitoring. Combining the Internet of Things technology with vegetable cleaning machines is expected to realize intelligent, efficient, and precise vegetable cleaning processes. SUMMARY

[0012] The purpose of the present application is to provide an intelligent vegetable cleaning machine monitoring system and method based on the Internet of Things to solve the problems raised in the background art.

[0013] To achieve the above-mentioned purpose, the present application provides an intelligent vegetable cleaning machine monitoring system based on the Internet of Things, which comprises: A device state monitoring module that collects real-time device state information of the vegetable cleaning machine, including power-on state, remote control mode, working state, and fault alarm state. A process parameter detection module is configured to monitor water temperature, water level, and turbidity in real time during the cleaning process. An industrial control module is configured to perform start-stop operation, sequence start-stop control, device interlock control, and emergency stop control of the vegetable cleaning machine based on the device state information and the process parameters. A data integration transmission module is configured to receive real-time data from the device state monitoring module and the process parameter detection module, and transmit the data to the intelligent analysis module through a wireless communication network. An intelligent analysis module is configured to establish a dynamic model of the vegetable cleaning process using the received data, and generate cleaning parameter optimization instructions. A cloud service module is configured to store historical data and real-time data, and provide remote access and monitoring functions.

[0014] Preferably, the device state monitoring module collects device state information through various sensors integrated in the vegetable cleaning machine, including using a current sensor to detect the power-on state of the motor, using a voltage sensor to monitor power stability, and using a vibration sensor to analyze the mechanical operation state of the device. The device state monitoring module has a signal conversion unit built-in to amplify and filter the analog signals output by the sensors and convert them into digital signals. The digital signals are analyzed in real time by an embedded microcontroller to extract device state characteristic values, including working period, fault frequency, and remote mode identifier. The analyzed data is packaged through a wireless communication protocol and transmitted to the industrial control module.

[0015] Preferably, the process parameter detection module uses multi-sensor fusion technology to monitor cleaning process parameters, including real-time measurement of water temperature using a platinum resistance temperature sensor, detection of water level using an ultrasonic water level sensor, and analysis of water turbidity using an optical turbidity sensor. The process parameter detection module is equipped with a data acquisition card to collect sensor data at a fixed sampling frequency and perform analog-to-digital conversion. The converted data is processed by a calibration algorithm to compensate for environmental errors and generate standardized parameter values. The parameter values are sent to the data integration transmission module through a serial communication interface.

[0016] Preferably, the industrial control module implements control logic based on a programmable logic controller architecture, including writing a ladder diagram program to define the device start-stop sequence, setting a timer to control the sequence start-stop timing, starting from the end device of the cleaning line and stopping from the head device. The industrial control module integrates a digital input-output module to receive fault signals from the device state monitoring module, trigger interlock control programs, and automatically stop upstream devices. The emergency stop circuit implements rapid power-off through a hardware relay. The control module also includes a communication interface to receive optimization instructions from the intelligent analysis module and dynamically adjust control parameters.

[0017] Preferably, the data integration transmission module adopts a gateway device to realize data aggregation and forwarding, including configuring a multi-protocol converter, supporting RS485 and LORA communication protocols, and parsing the data packets of the device state monitoring module and the process parameter detection module; the data integration transmission module runs a data compression algorithm to reduce the transmission bandwidth, and adds a timestamp and a data check code; the compressed data is transmitted to the intelligent analysis module through the TCP / IP protocol to ensure data integrity and real-time performance; the gateway device also has a cache mechanism to temporarily store data when the network is interrupted.

[0018] Preferably, the intelligent analysis module uses a deep learning algorithm to establish a cleaning process dynamic model, including collecting historical device state and process parameter data as a training set, using a convolutional neural network to extract features, and predicting cleaning efficiency trends; the intelligent analysis module receives streaming data from the data integration transmission module in real time, calculates optimal water temperature, water level and turbidity set values through an inference engine; the optimization instruction generation module iteratively adjusts parameters based on the prediction results using the gradient descent method, and outputs control instructions to the industrial control module.

[0019] Preferably, the cloud service module is deployed on a cloud server to realize data storage and remote access, including configuring a relational database to store device state records, process parameter history and alarm logs; the cloud service module provides a RESTful API interface to allow remote users to query real-time data and generate reports; the data synchronization module periodically pulls optimization records from the intelligent analysis module for backup and version management; the cloud service module also integrates identity authentication and encrypted transmission to ensure data security.

[0020] Preferably, the system further includes a local display module to realize data visualization through an embedded touch screen, including designing a graphical user interface to display device state graphs, process parameter curves and alarm lists; the local display module connects the industrial control module and the intelligent analysis module to refresh data in real time; the data export function allows users to save historical data through a USB interface; the touch screen driver supports gesture operations for parameter setting and instruction input.

[0021] Preferably, an adaptive feedback control loop is established between the industrial control module and the intelligent analysis module, including the industrial control module sending device running status to the intelligent analysis module in real time, the intelligent analysis module adjusting cleaning model parameters according to the received data to generate new optimization instructions; the feedback loop uses a PID control algorithm to dynamically correct control bias to ensure system stability; the communication delay compensation module estimates transmission delay to synchronize control instructions.

[0022] Preferably, the application further includes an intelligent vegetable cleaning machine monitoring method based on the Internet of Things, including all modules and method processes of the above-mentioned intelligent vegetable cleaning machine monitoring system based on the Internet of Things.

[0023] Compared with the prior art, the present application has the following advantages: The device state monitoring module can collect information such as the power-on state, remote control mode, working state, and fault alarm state of the vegetable cleaning machine in real time. Users do not need to be near the cleaning machine at all times, and can understand the running status of the cleaning machine at any time through the connected terminal device. Once a fault occurs in the device, such as damage to a certain part causing abnormal working state, the fault alarm state will immediately issue an alarm to notify the maintenance personnel. This enables the maintenance personnel to respond quickly and promptly repair the device, avoiding production interruptions caused by device failures, greatly improving the availability and production efficiency of the device, and reducing maintenance costs. At the same time, real-time monitoring of the power-on state and remote control mode facilitates management personnel to flexibly adjust the running mode of the device according to actual production needs, realizing fine management of the device.

[0024] The process parameter detection module monitors the water temperature, water level, and turbidity parameters in real time during the cleaning process, providing precise and suitable conditions for vegetable cleaning. Different types of vegetables have different requirements for water temperature during cleaning. For example, leafy vegetables are suitable for cleaning at lower water temperatures to avoid damage to the leaves, while root vegetables may require slightly higher water temperatures to better remove surface dirt. Through real-time monitoring and precise control of water temperature by this module, the most suitable water temperature can be set according to the type of vegetables, thereby ensuring the cleaning effect while maximizing the preservation of the nutritional content and taste of the vegetables.

[0025] Water level monitoring is also important. A suitable water level can ensure that the vegetables are fully soaked and cleaned, and can also avoid waste of water resources. When the water level is too low, the module will promptly feedback and prompt the operator to add water to ensure the smooth progress of the cleaning process.

[0026] The turbidity parameter reflects the degree of pollution of the cleaning water. When the turbidity of the cleaning water reaches a certain value, it means that there are too many impurities in the water, and continued use will affect the cleaning effect. At this time, the system can automatically prompt to replace the cleaning water according to the turbidity monitoring results, ensuring that each cleaning is carried out in a clean water environment, thereby ensuring the stability of the cleaning quality.

[0027] The industrial control module performs various control functions such as start-stop operation, sequential start-stop control, device interlocking control, and emergency stop control of the vegetable cleaning machine based on device state information and process parameters. In terms of start-stop operation, operators can easily issue instructions through a remote terminal without needing to personally operate at the device site, saving time and manpower. The sequential start-stop control function can start or stop each component of the cleaning machine in sequence according to the preset program, avoiding damage to the device due to improper operation. For example, when starting, the water supply system is started first, and the cleaning motor is started after the water level reaches a certain height, ensuring the normal operation of the device.

[0028] Device interlock control enhances the safety and reliability of equipment operation. When a critical component fails, other components associated with it will automatically stop running to prevent the failure from spreading. For example, when the cleaning motor overheats, the device interlock control will immediately stop the water pump from running, preventing more serious damage from lack of water cooling.

[0029] Emergency stop control plays a key role in the face of sudden emergencies. Once the emergency stop button is triggered, the system will respond quickly and immediately stop all running components of the cleaning machine, minimizing the potential loss and harm, and ensuring the safety of personnel and equipment. This intelligent control method makes the operation of the cleaning machine more flexible and efficient, and can adapt to different production scenes and needs.

[0030] The data integration transmission module receives real-time data from the equipment state monitoring module and the process parameter detection module, and sends the data to the intelligent analysis module through a wireless communication network. The intelligent analysis module uses these rich data to establish a dynamic model of the vegetable cleaning process. Through analysis and research of the model, the relationships between various factors in the cleaning process can be understood in depth, such as the correlation between water temperature, cleaning time, cleaning intensity and cleaning effect.

[0031] The intelligent analysis module can generate cleaning parameter optimization instructions. For example, through analysis of historical data, it is found that when cleaning a certain type of vegetable, increasing the water temperature by a certain number of degrees and appropriately extending the cleaning time can significantly improve the cleaning effect while reducing energy consumption during the cleaning process. The system will automatically adjust the cleaning parameters based on these analysis results to optimize the cleaning process, improve cleaning efficiency and quality, and reduce production costs. Moreover, the integration and analysis of data also provide strong support for the production decisions of enterprises, which can reasonably arrange production plans and optimize resource allocation based on data analysis results. BRIEF DESCRIPTION OF DRAWINGS

[0032] Figure 1 The working principle diagram of the intelligent vegetable cleaning machine monitoring system based on the Internet of Things described in the present application; Figure 2 The flowchart of data acquisition and transmission of the equipment state monitoring module; Figure 3 The flowchart of data monitoring and transmission of the process parameter detection module; Figure 4 The running state diagram of the industrial control module; Figure 5 The intelligent data analysis optimization effect diagram. DETAILED DESCRIPTION

[0033] With reference to the drawings of the embodiments of the present application, the technical solutions in the embodiments of the present application will be described clearly and completely. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments of the present application, all the other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of the present application.

[0034] Please refer to Figure 1 The present application provides a kind of based on Internet of Things's intelligent vegetable cleaning machine monitoring system and method, the system includes: equipment state monitoring module, process parameter detection module, industrial control module, data integration transmission module, intelligent analysis module and cloud service module.Specific implementation is as follows: Equipment state monitoring module acquires the equipment state information of vegetable cleaning machine in real time, equipment state information covers power-on state, remote control mode, work running state and fault alarm state.Process parameter detection module monitors water temperature, water level and turbidity parameter in real time in cleaning process.Industrial control module is based on equipment state information and process parameter and executes the start-stop operation of vegetable cleaning machine, sequence start-stop control, equipment interlock control and emergency stop control.Data integration transmission module receives the real-time data of equipment state monitoring module and process parameter detection module, and data is sent to intelligent analysis module by wireless communication network.Intelligent analysis module establishes vegetable cleaning process dynamic model using received data, generates cleaning parameter optimization instruction.Cloud service module stores historical data and real-time data, provides remote access and monitoring function.

[0035] Example 1: refer to Figure 2The device state monitoring module realizes the collection of device state information through various sensors integrated on the vegetable cleaning machine. The device state information includes power-on state, remote control mode, working state and fault alarm state. The sensor network of the device state monitoring module includes current sensor, voltage sensor and vibration sensor. The current sensor is directly clamped on the three-phase power supply line of the vegetable cleaning machine drive motor, which detects the on-off state and current intensity change in real time. The output signal of the current sensor is directly related to the power-on state of the motor. When the motor is powered on, the current signal presents a stable sinusoidal waveform, and when the motor is powered off, the current signal is zero. The voltage sensor is connected in parallel to the main power input of the vegetable cleaning machine to monitor the voltage stability of the power supply network. The voltage sensor can capture voltage transient fluctuations, surges or drop phenomena. The unstable characteristics of the voltage signal are associated with power supply failure. The vibration sensor is fixedly installed on the drum bearing seat of the vegetable cleaning machine through a magnetic base to collect vibration acceleration signals generated by the mechanical operation of the device. The frequency spectrum characteristics of the vibration signal reflect the mechanical operation state such as bearing wear and rotor imbalance. The signal conversion unit built-in the device state monitoring module pre-processes the original analog signals output by the sensor. The signal conversion unit includes multi-stage operational amplifier circuit and active filter circuit. The current signal or voltage signal output by the current sensor and voltage sensor has low amplitude. The operational amplifier circuit amplifies the weak analog signal with fixed gain. The amplification factor is set according to the sensor range and the input range of the subsequent analog-to-digital converter. The signal voltage range after amplification matches the reference voltage of the analog-to-digital converter. The active filter circuit adopts Butterworth low-pass filter topology structure, and the cutoff frequency is set to twice the upper limit of the useful frequency of the signal. The filter effectively suppresses high-frequency electromagnetic interference and power harmonic noise, and the smoothed analog signal waveform retains the main characteristics of the original signal. The analog-to-digital converter converts the pre-processed analog signal into digital signal. The analog-to-digital converter adopts successive approximation architecture with 16-bit resolution and sampling rate of 1000 times per second. The analog-to-digital conversion process quantizes and encodes each analog sampling point to generate a binary digital sequence.

[0036] The embedded microcontroller receives the digital signal output by the analog-to-digital converter and performs real-time analysis. The embedded microcontroller is based on the ARM Cortex-M4 core and runs a real-time operating system. The analysis algorithm extracts device state characteristic values from the digital signal sequence, including the working period, fault frequency, and remote mode identifier. The working period calculation is based on the duration statistics of the motor current signal. A complete working period starts from the rising edge of the current and ends at the falling edge. The fault frequency is obtained by analyzing the number of voltage abnormal events and vibration overrun events. If the fault event count per unit time exceeds the threshold, a fault alarm state is triggered. The remote mode identifier comes from the digital input signal of the vegetable washing machine control panel. In the remote control mode, the digital input is high. The embedded microcontroller stores the threshold range of the device state characteristic values. When the characteristic values exceed the threshold range, the device state information is updated. The data transmission of the device state monitoring module uses a wireless communication protocol. The wireless communication protocol selects LoRa spread spectrum technology. The embedded microcontroller encapsulates the device state information obtained by analysis into a data packet. The data packet structure includes a preamble, a frame header, a device address, a data payload, and a cyclic redundancy check code. The data payload part encodes the specific content of the device state information. The power-on state is represented by a binary number. The working state is encoded by two binary numbers to represent different running modes. The fault alarm state occupies one byte to store the fault code. The cyclic redundancy check code is used for data packet transmission error detection. The receiving party can request retransmission if the check fails. The transmission power of the wireless communication module is adjustable. The communication distance covers the workshop where the vegetable washing machine is located. The data packet sending interval is dynamically adjusted according to the device state change rate. When the state is stable, the sending period is lengthened. When the state changes frequently, the sending period is shortened.

[0037] The mechanical structure of the device condition monitoring module meets the requirements of industrial environment application, and the protection level of the sensor shell reaches the IP67 standard to prevent water vapor and impurities from entering during cleaning. The signal conversion unit and embedded microcontroller are installed in a sealed electrical control box, which is equipped with a temperature control device to maintain the electronic components within the appropriate temperature range. The power management circuit provides stable DC power for the device condition monitoring module, and has overvoltage protection and reverse connection protection functions. The overall layout of the device condition monitoring module considers electromagnetic compatibility, with separate signal and power lines, and separate analog and digital circuit layouts to reduce mutual interference. The device condition monitoring module requires initial configuration and calibration, with configuration parameters including sensor range, communication address, and sampling frequency. Calibration is performed after installation, and the calibration personnel connect the device condition monitoring module through the debugging interface using a special configuration tool to set the current variable ratio, voltage multiplier, and vibration sensitivity parameters. The calibration data is stored in the non-volatile memory of the embedded microcontroller, and the configuration parameters are not lost after power failure. The device condition monitoring module supports remote diagnosis and firmware upgrade, and the maintenance personnel access the internal state of the device condition monitoring module through network connection to upload new version firmware program to improve functionality. The interface between the device condition monitoring module and the industrial control module uses optical isolation technology, and the digital output signal is transmitted through an optocoupler to achieve electrical isolation. The isolation design prevents high-voltage signals from the industrial site from entering the low-voltage circuit of the device condition monitoring module, protecting the core electronic components from damage. The device condition monitoring module has a fault self-diagnosis function that periodically checks the working state of internal sensors and circuits, including power voltage detection, memory verification, and communication link testing. When the self-diagnosis is abnormal, the device condition monitoring module sends its fault signal to the industrial control module, requesting maintenance intervention.

[0038] The long-term operation of the device state monitoring module generates a large amount of device state historical data, which is stored in the external storage chip of the embedded microcontroller. The storage chip uses a serial Flash memory type, and the storage capacity meets the needs of continuous recording of 30 days of data. The historical data record format includes a timestamp, device state characteristic values, and environmental temperature information, the timestamp is provided by a hardware clock circuit, and the time synchronization is achieved through the timing broadcast of the industrial control module. Maintenance personnel can periodically export historical data for analyzing device operation trends, and data export is completed through a wireless communication protocol or a physical interface. The low-power design of the device state monitoring module prolongs the working time of the device under standby power supply, the embedded microcontroller enters sleep mode during idle periods, and the wireless communication module switches to a low-power state during data transmission gaps. The power management circuit monitors the battery power, and sends a warning signal when the power is insufficient. The installation bracket of the device state monitoring module is made of stainless steel, and the corrosion resistance is suitable for high-humidity cleaning environments. The sensor cable connector uses a waterproof type, and the plug-in frequency meets the needs of frequent maintenance. The surface identification of the device state monitoring module is clear, and the identification content includes device number, model specification, and electrical parameters. The software architecture of the device state monitoring module includes a bottom layer driver, middleware, and an application program. The bottom layer driver directly operates hardware peripherals, including an analog-to-digital converter driver, a serial port driver, and a timer driver. The middleware layer implements a communication protocol stack and data cache management, the communication protocol stack processes packet assembly and disassembly processes, and the data cache management uses a ring buffer structure. The application program layer executes the main signal analysis and state judgment logic, and the application program uses modular design, and the function modules exchange data through a message queue. The software code is written in C language, and the code structure is clear and convenient for subsequent function extension and maintenance modification. The environmental adaptability of the device state monitoring module is strictly tested, and the test items include high temperature and high humidity test, vibration and impact test, and electromagnetic interference test. The high temperature and high humidity test simulates the high temperature and high humidity environment of the cleaning workshop, and the device state monitoring module works continuously for 100 hours under the condition of temperature 85 degrees Celsius and relative humidity 95%. The vibration and impact test simulates the mechanical vibration during device operation, the vibration frequency range covers 5 Hz to 2000 Hz, and the acceleration reaches 5 times the gravity acceleration. The electromagnetic interference test applies high frequency radiation field strength and fast transient pulse group to verify the anti-interference ability of the device state monitoring module.

[0039] Example 2: see Figure 3The process parameter detection module uses a multi-sensor fusion technology to monitor the cleaning process parameters, including water temperature, water level and turbidity. A platinum resistance temperature sensor is installed inside the water tank of the vegetable cleaning machine, with the probe of the platinum resistance temperature sensor directly immersed in the cleaning water. The resistance value of the platinum resistance shows a good linear relationship with temperature changes. The measurement circuit uses a four-wire connection method to eliminate lead resistance errors. A constant current source provides a stable excitation current, and a voltage measuring instrument detects the voltage drop across the platinum resistance. According to the platinum resistance scale, the resistance value is converted to a temperature value. An ultrasonic water level sensor is fixedly installed on the top of the water tank. The ultrasonic transducer emits high-frequency sound wave pulses facing the water surface. When the sound wave encounters the water surface, it reflects to form a return wave. The computing unit inside the ultrasonic water level sensor measures the time difference between transmission and reception, and calculates the water level height in combination with the propagation speed of sound waves in air. The installation position of the sensor avoids the turbulent area of the water inlet and the drain outlet, ensuring a smooth liquid surface for measurement. The optical turbidity sensor uses a transmission light measurement principle. The light source of the optical turbidity sensor emits infrared light of a specific wavelength. After the light beam passes through the water sample, it reaches the photodetector on the opposite side. The photodetector converts the light intensity signal into a current signal. The suspended particulate matter in the water sample scatters and absorbs light, and the transmission light intensity is inversely proportional to the turbidity value. The data acquisition card provided by the process parameter detection module is responsible for collecting the output signals of each sensor. The data acquisition card integrates multiple analog input channels and an analog-to-digital converter. The output signal of the platinum resistance temperature sensor is a millivolt-level voltage signal. The analog input channel of the data acquisition card is configured with a programmable gain amplifier to amplify the small signal to a voltage range suitable for the input of the analog-to-digital converter. The ultrasonic water level sensor and the optical turbidity sensor output standard current signals. The input channel of the data acquisition card is equipped with a precision sampling resistor to convert the current signal to a voltage signal. The analog-to-digital converter discretely samples the analog signal at a fixed sampling frequency. The sampling frequency is set to 100 Hz according to the characteristics of the sensor signal and the system control requirements. The resolution of the analog-to-digital converter is 16 bits, with 65536 levels of quantization. The conversion speed of the analog-to-digital converter meets the timing requirements of the multi-channel round-robin sampling. Each input channel of the data acquisition card is equipped with an independent sample-and-hold circuit to ensure the synchronization of multi-channel signal sampling. The sample-and-hold circuit captures the instantaneous value of the signal at the sampling time and maintains the voltage stable during the conversion period.

[0040] The converted digital data is processed by a calibration algorithm stored in the microprocessor of the process parameter detection module. The calibration of the platinum resistance temperature sensor uses a polynomial fitting method, the output voltage of the platinum resistance is measured at multiple known temperature points, a temperature and voltage correspondence table is established, and the accurate temperature value is obtained by table lookup interpolation during actual measurement. The calibration of the ultrasonic water level sensor compensates for the influence of environmental temperature on sound velocity, the formula for the relationship between sound velocity and temperature change is stored in the memory, and the real-time temperature value is measured by an additional temperature probe. The calibration of the optical turbidity sensor uses a standard turbidity liquid to calibrate, a linear regression equation of turbidity value and output current is established, and periodic calibration is used to eliminate errors introduced by light source aging and detector sensitivity changes. The calibrated data is converted into standard engineering units, the water temperature unit is Celsius, the water level unit is millimeter, and the turbidity unit is NTU, and the converted parameter value is stored in the data buffer area. The process parameter detection module is connected with the data integration transmission module through a serial communication interface, and the serial communication interface uses the RS485 standard. The data frame format includes a start bit, a data bit, a check bit and a stop bit, the communication baud rate is set to 9600bps, the data bit length is 8 bits, the check mode is even check, and the stop bit length is 1 bit. Each data frame transmits the measurement value of one parameter point, and the data frame header includes sensor address coding to distinguish the data sources of different sensors. The communication protocol uses a master-slave question and answer mechanism, the data integration transmission module is used as the master station to poll the process parameter detection module regularly, and the process parameter detection module is used as the slave station to respond to the data request. A twisted shielded cable is used for the communication cable to reduce the influence of electromagnetic interference on signal transmission quality, and a terminal resistor matches the cable characteristic impedance to suppress signal reflection.

[0041] The mechanical structure design of the process parameter detection module considers the special requirements of the cleaning environment, the sensor probe material is selected as corrosion-resistant stainless steel 316L type, the sealing structure reaches IP68 protection level, and the long-term immersion does not leak. The protective sleeve of the platinum resistance temperature sensor adopts thin-wall design, reduces thermal inertia and improves response speed, and the installation position is close to the middle of the water tank to avoid direct radiation of the heating element. The surface of the ultrasonic water level sensor is coated with anti-condensation material to prevent water vapor condensation from affecting the efficiency of sound wave emission. The measurement cavity of the optical turbidity sensor is designed with self-cleaning function, and high-pressure water flow is periodically introduced to flush the mirror surface to prevent dirt from adhering. The data acquisition card is installed in a waterproof junction box, the inlet and outlet ports of the junction box use waterproof joints, and desiccant is placed in the box to control humidity. The power supply of the process parameter detection module uses a safe voltage of 24V DC, and the power supply module has overcurrent protection and short circuit protection functions. The internal circuit of the module is divided into analog circuit and digital circuit two areas, the analog circuit area includes sensor signal conditioning circuit, the digital circuit area includes microprocessor and communication interface, the two areas are electrically isolated by photoelectric isolator, to prevent digital circuit noise from interfering with analog signals. The power supply wiring and signal wiring are separated, the analog ground and digital ground are connected at a single point, to reduce ground loop interference. The circuit board is coated with three-proof paint for protection, to enhance moisture and corrosion resistance. The software program of the process parameter detection module realizes the coordination of data acquisition, processing and communication, and the software program is developed based on a real-time operating system. The data acquisition task is triggered by a timing interrupt, the interrupt service program starts the analog-to-digital converter for conversion, and the conversion result is read after conversion. The data processing task performs digital filtering on the original data, uses a moving average algorithm to smooth random interference, and adjusts the filter window size according to the signal characteristics. The communication task manages the data transmission of the RS485 interface, receives data and integrates the query command of the transmission module, organizes the response data frame and sends it. The software program sets a watchdog timer, which automatically resets the system to resume operation when the program runs away. The parameter setting function allows the running parameters such as sampling frequency and communication address to be modified through the configuration tool, and the configuration data is stored in non-volatile memory.

[0042] The installation and debugging of the process parameter detection module requires professional personnel to select representative measurement points at the installation location, avoiding dead water zones and turbulent flow zones. The insertion depth of the platinum resistance temperature sensor reaches more than two-thirds of the pipe diameter, ensuring that the main fluid temperature is measured. The installation of the ultrasonic water level sensor ensures that the transducer surface is parallel to the liquid surface, and the measurement reference surface is aligned with the bottom of the water tank. The installation of the optical turbidity sensor reserves straight pipe sections before and after to ensure smooth water flow through the measurement cavity. During the debugging process, standard instruments are used for comparative measurement, and the calibration parameters are adjusted to make the display value consistent with the actual value. After debugging is completed, the adjustable parameters are locked to prevent misoperation from changing the settings. Maintenance work includes regular calibration and cleaning, and the calibration period is determined according to the frequency of use and environmental conditions, generally not more than six months. Calibration uses standard instruments that have been measured and verified, and calibration records are archived for reference. The sensor probe is periodically disassembled and cleaned to remove surface deposits and restore measurement accuracy. The light source and detector of the optical turbidity sensor have a lifespan of about 10,000 hours, and need to be replaced when they approach the end of their lifespan. Maintenance personnel read the module operating status through the diagnostic interface, and the diagnostic information includes power voltage, communication error count, and self-test results, which are handled in a timely manner. The electromagnetic compatibility design meets the standards for industrial environment applications, and the circuit board layout ensures that the area of high-frequency signal loops is minimized, and key signal lines are equipped with magnetic beads to suppress high-frequency noise. A pressure-sensitive resistor and transient suppression diode are installed at the power input end to absorb lightning surges and operating overvoltages. The housing is made of metal material and is well grounded, providing electromagnetic shielding. The module passes electromagnetic compatibility tests such as electrostatic discharge and radiation immunity to ensure stable operation in complex industrial environments. The data output format is standardized, and each parameter value is accompanied by a quality code to identify the reliability of the data. The quality code is determined based on the sensor status, signal strength, and calibration status. The data timestamp is provided by a hardware clock, and the clock accuracy reaches an error of less than one second per day. The historical data cache function temporarily stores the last 1000 groups of data when communication is interrupted, and the missing data is transmitted after communication is restored.

[0043] In Example 3, the industrial control module implements control logic based on a programmable logic controller architecture. The central processing unit of the programmable logic controller uses a multi-core processor structure with a main frequency of 1.2 GHz, and is equipped with a real-time operating system to ensure deterministic response of control tasks. The ladder programming environment provides a rich library of function blocks, including timer function blocks, counter function blocks, and comparator function blocks. The device start-stop sequence is implemented through the network structure of the ladder program, with each network containing a logical combination of normally open contacts, normally closed contacts, and output coils. The sequential start-stop timing is controlled by a timer function block, which sets a hold-on delay timer. The timer starts from the end device of the cleaning line and stops the head device, with the time interval set according to process requirements. The digital input and output module uses photoelectric isolation technology. The input module receives fault signals from the device state monitoring module, and the output module drives the contactor to control the motor start-stop. The input and output points are reserved with a 20% excess to meet expansion needs. The interlock control program of the industrial control module is triggered based on fault signals. The interlock control program uses a priority interrupt mechanism. Upon input of a fault signal, the current main program scanning period is immediately suspended, and the interlock control program is executed. The interlock control program traverses the device dependency table to determine the upstream device number that needs to be stopped, sends a stop command through the digital output module, and resumes the main program after the interlock action is completed. The emergency stop circuit is independent of the programmable logic controller software. The emergency stop circuit uses hardware relays to build a safety circuit. The emergency stop button is connected in series with normally closed contacts. When the emergency stop button is actuated, the control circuit power is cut off, achieving a fast power-off with a safety level of PLd. The communication interface of the control module supports multiple industrial protocols. The communication interface module integrates an Ethernet port and a serial port. The Ethernet port uses the PROFINET protocol to communicate with the intelligent analysis module, and the serial port uses the Modbus RTU protocol to connect to the human-machine interface.

[0044] The dynamic parameter adjustment function of the industrial control module receives the optimization instruction of the intelligent analysis module, and the optimization instruction is transmitted to the data register of the programmable logic controller through the communication interface. The data register stores parameters such as water temperature set value, water level set value and turbidity threshold value. The user program of the programmable logic controller periodically reads the data register, applies the new parameters to the PID control algorithm, and the output of the PID control algorithm adjusts the heater power and the water valve opening. The communication delay compensation module estimates the network transmission delay, and the delay estimation value is used to adjust the execution time of the control instruction to maintain the accuracy of the control timing. The power supply design of the industrial control module adopts a redundant structure, and the main power module and the standby power module are connected in parallel to supply power. The power module has overvoltage protection and overcurrent protection functions. When the main power fails, it automatically switches to the standby power, and the switching time is less than 10 milliseconds, ensuring the continuous operation of the control module. The bottom bus of the programmable logic controller adopts parallel communication mode, and the transmission speed of the bottom bus reaches 100Mbps, ensuring the real-time data exchange between the central processing unit and the input / output module. The module status indicator displays the running state, fault state and communication state, which is convenient for field maintenance personnel to quickly diagnose. The software program structure includes initialization program, main loop program and interrupt service program. The initialization program is executed when the programmable logic controller is powered on, which sets the hardware parameters, clears the data area and establishes the communication connection. The main loop program periodically scans the ladder logic, and the scan period of the main loop program is fixed at 10 milliseconds. Each scan period completes three stages of input sampling, program execution and output refresh. The interrupt service program processes high-priority events, and the interrupt sources include emergency stop signal, communication interruption and hardware fault alarm. Program variables use symbolic addressing, and variable names are bound to physical addresses to improve program readability and maintainability. The installation requirements meet the electrical control cabinet specifications, and the control module is installed on the standard guide rail with reserved space around for heat dissipation. The wiring uses crimp terminals, and the wire number is clearly visible. The power lines and signal lines are laid separately to reduce interference.

[0045] The debugging tool of the industrial control module provides online monitoring functions. The debugging tool connects the programming port to display variable values in real time, forcibly sets input and output states, and tests external devices. The program simulation function verifies the logic correctness without hardware connection, and the breakpoint debugging function executes the program step by step to locate errors. After debugging, the program code is encrypted for protection against unauthorized modification. Version management records program modification history, and each modification saves backup files. Maintenance work includes regular checks and work status records. Regular check items include power voltage measurement, terminal tightening, and cooling fan cleaning. Work status records are stored in non-volatile memory, and the records contain running time, fault frequency, and communication error statistics. Preventive maintenance prompts component replacement based on device running time, such as replacing the backup battery every three years to ensure that clock data is not lost during power failure. The adaptive feedback control loop establishes a closed-loop control system, and the industrial control module sends device running state data packets to the intelligent analysis module in real time. The data packets include motor current, valve position, and actual temperature values. The intelligent analysis module adjusts the cleaning model parameters based on the received data and generates new optimization instructions for the industrial control module. The feedback loop uses an incremental PID control algorithm to dynamically correct control bias. The parameters of the PID control algorithm are determined using the critical proportionality method to determine the proportionality coefficient, integral time, and differential time. The stability analysis of the control system is based on the Routh criterion, and all roots of the characteristic equation are located in the left half-plane to ensure system stability.

[0046] Referring to Figure 4 , the running state monitoring data of the industrial control module based on the programmable logic controller architecture in the vegetable cleaning process is shown. The chart contains the real-time variation trends of multiple key parameters such as motor current, heater power, water valve opening, and control accuracy. The motor current curve reflects the load situation of the cleaning device in different working stages. The current remains relatively stable during normal operation, and fluctuations may correspond to device start-stop or load changes. The heater power data shows the response characteristics of the temperature control system, and power adjustment directly affects the stability of the cleaning water temperature. The water valve opening parameter reflects the flow control accuracy, and the opening change is closely related to water level regulation. The control accuracy curve shows the overall control performance of the system, and a high accuracy value indicates stable system operation and accurate parameter adjustment. The fault event points marked in the figure correspond to abnormal conditions detected by the device state monitoring module. These points trigger the interlock control program to automatically execute the corresponding protection measures. Through multi-parameter collaborative analysis, the running efficiency of the control module can be comprehensively evaluated, potential problems can be found in time, and the stability and reliability of the vegetable cleaning process can be ensured.

[0047] In embodiment 4, the data integration transmission module is implemented by using a gateway device to realize data aggregation and forwarding. The hardware platform of the gateway device is based on an ARM architecture processor with a main frequency of 1.5 GHz, equipped with 1 GB of running memory and 8 GB of embedded storage. The multi-protocol converter is the core component of the gateway device, supporting the analysis and conversion of RS485 and LORA communication protocols. The RS485 interface circuit uses a differential signal transmission method, which has strong anti-common mode interference ability, and the communication baud rate can be configured in the range of 9600 bps to 115200 bps. The LORA wireless module uses spread spectrum modulation technology, with a receiving sensitivity of -137 dBm. The gateway device periodically polls the process parameter detection module connected to the RS485 bus and simultaneously listens to the LORA wireless channel to receive data packets from the device state monitoring module. The data packet analysis engine disassembles the data frame according to the predefined protocol rules, extracts the valid payload data, and the payload data includes sensor readings, device status codes, and timestamp information.

[0048] The data integration transmission module runs a data compression algorithm to reduce network transmission bandwidth occupation. The data compression algorithm uses the LZ77 algorithm based on a dictionary, with a sliding window size of 32 KB and a look-ahead buffer size of 256 bytes. The compression process identifies repeated sequences in the data stream and replaces them with length-distance pairs. The compressed data block is added with a frame header marker and a CRC32 check code. The timestamp data is provided by a high-precision real-time clock chip in the gateway device, which is synchronized with the network time server through the NTP protocol to ensure that all data has a unified time reference. The data check code uses the cyclic redundancy check algorithm, with a check polynomial of CRC-16-CCITT, to ensure the integrity of the data during transmission. The compressed and encapsulated data is transmitted to the intelligent analysis module through the TCP / IP protocol. The Ethernet interface of the gateway device supports the 1000BASE-T standard and uses an RJ45 connector. The TCP socket connection maintains a long connection state to reduce the overhead of connection establishment. The gateway device also has a cache mechanism, using an embedded multimedia card with a storage capacity of 8 GB. When the network is interrupted or the intelligent analysis module is unavailable, data is temporarily written to the cache area. After the network is restored, the data is retransmitted in chronological order.

[0049] The intelligent analysis module is deployed on an edge computing server, which is configured with a quad-core processor, 16 GB of memory, and a 256 GB solid state drive. The intelligent analysis module uses a deep learning algorithm to establish a dynamic model of the cleaning process. The deep learning framework uses TensorFlow2.0, and the historical data set required for model training is downloaded in bulk from the cloud service module. The historical data set includes device state records, process parameter history, and control instruction logs. The data preprocessing process includes data cleaning, normalization, and feature engineering. The data cleaning step removes records containing obvious errors and duplicate entries. Normalization scales sensor data of different dimensions to the [0, 1] interval. Feature engineering constructs time-domain statistical features such as moving average, standard deviation, and slope. The convolutional neural network model includes three convolutional layers, two max-pooling layers, and one fully connected layer. The convolutional layers use ReLU activation functions, and the convolutional kernel size is 3x3. The pooling layer window size is 2x2. The model training uses the Adam optimizer, the loss function is mean squared error, the training iteration number is set to 1000 rounds, and the batch size is 128. The trained convolutional neural network model is deployed in the inference engine, which uses the TensorFlow Serving component to provide gRPC interface services. The real-time streaming data of the data integration transmission module is continuously input to the inference engine through the gRPC interface. The inference engine loads the convolutional neural network model to perform forward propagation calculation on the input data, and outputs the predicted value of the cleaning efficiency. The optimization instruction generation module iteratively adjusts the control parameters based on the prediction results using the gradient descent method. The learning rate of the gradient descent method is set to 0.001, and the momentum parameter is set to 0.9. The iteration process aims to minimize the difference between the predicted cleaning efficiency and the target efficiency. The optimal water temperature set value, water level set value, and turbidity set value calculated by the optimization instruction generation module are packaged into a JSON format instruction message, which is published to a topic through the MQTT protocol. The industrial control module subscribes to the topic to receive optimization instructions.

[0050] The data flow between the data integration transmission module and the intelligent analysis module is configured with quality monitoring. The data quality monitoring module checks the timeliness, integrity, and reasonableness of the data. The timeliness check determines whether the difference between the data timestamp and the current time exceeds the threshold, the integrity check verifies whether the data frame length conforms to the protocol specification, and the reasonableness check evaluates whether the sensor values are within the physically possible range. Abnormal data is marked and recorded in the log file, and does not participate in model inference and optimization calculation. The communication link between the two modules implements bidirectional authentication and encrypted transmission. Bidirectional authentication uses X.509 digital certificates, and encrypted transmission uses the TLS1.3 protocol to prevent data leakage and tampering.

[0051] The key parameters involved in the system operation process are stored in a structured configuration file. The configuration file is in YAML format, which is easy to read and modify. The following table lists the main configuration parameters of the data integration transmission module and the intelligent analysis module interface and their meanings: Table 1: Data integration transmission module and intelligent analysis module interface configuration parameter table The model updating mechanism of the intelligent analysis module supports online learning. The online learning process periodically uses the latest operation data to perform incremental training on the convolutional neural network model. The incremental training period is set to 24 hours, and the training process is performed during periods of low system load. The model version management tool tracks the performance of different versions of the model, and the performance evaluation indicators include prediction accuracy and root mean square error. When a new version of the model performs better than the old version on the validation set, it is automatically switched to the new model. System administrators can manually trigger model training and model rollback operations through the management interface. The management interface is developed based on Web technology and provides a graphical model performance monitoring dashboard. The firmware of the data integration transmission module and the software of the intelligent analysis module both support remote upgrading. The upgrade package is downloaded from the deployment server through a secure file transfer protocol. The upgrade process includes signature verification and integrity checking to prevent the execution of malicious code. The upgrade operation supports breakpoint resume and version rollback functions, and the upgrade log records the upgrade time, version number, and operation result in detail. The system generates a running status report periodically, which contains data transmission statistics, model inference times, and optimization instruction history. The report is generated in PDF format and can be automatically sent to designated administrators by email.

[0052] Referring to Figure 5 , the optimization effect of the data integration transmission module and the intelligent analysis module working together is presented. A double-Y-axis design is used, with the left side showing the cleaning efficiency and model accuracy, and the right side showing the optimization effect and data transmission volume. The cleaning efficiency curve reflects the optimization effect of the deep learning model on the cleaning process. The upward trend of the efficiency value indicates that the optimization instructions generated by the intelligent analysis module effectively improve the cleaning quality. The model accuracy data reflects the performance of the convolutional neural network in predicting cleaning efficiency. High accuracy ensures the scientificity of the optimization instructions. The optimization effect curve shows the actual effectiveness of the gradient descent algorithm in adjusting the control parameters. A positive value indicates the degree of improvement compared to the baseline state. The data transmission volume indicator reflects the data processing capacity of the gateway device. Stable data transmission is the basis for ensuring the real-time performance of the system. The synergistic relationship between the parameters reveals the working mechanism of the intelligent analysis module: the model continuously optimizes the control parameters based on real-time data, and continuously improves the system performance through a feedback control loop. This data-driven optimization method significantly improves the intelligence level of the vegetable cleaning process.

[0053] In the embodiment 5, the cloud service module is deployed on a cloud server cluster, the cloud servers use virtualization technology to achieve dynamic allocation of resources, and a load balancer distributes external access requests to multiple application server instances. The relational database selects MySQL 8.0 version, and the database table structure design includes a device state record table, a process parameter history table, and an alarm log table. The device state record table stores the power-on state, remote control mode, working state, and fault alarm state of the vegetable washing machine, and the fields include device number, state type, state value, and timestamp. The process parameter history table records the time series data of water temperature, water level, and turbidity, and each data point is associated with a device number and collection time. The alarm log table stores detailed information of fault events, including alarm code, alarm level, trigger time, confirmation time, and processing measures. Database indexes are established on frequently queried fields, such as device number and timestamp, and query performance is optimized through execution plan analysis tools.

[0054] The local display module realizes data visualization through an embedded touch screen. The hardware specifications of the embedded touch screen are a 10.1-inch IPS liquid crystal screen with a resolution of 1280x800 and support for 10-point capacitive touch. The graphical user interface is developed using the Qt framework, and the main interface layout is divided into three areas. The top area displays a device status diagram, which uses vector graphics to depict a simplified schematic diagram of the vegetable washing machine. Key components such as water pumps, conveyors, and sprinkler devices are colored according to their actual operating state, with green indicating operation, gray indicating stop, and red indicating failure. The middle area draws process parameter curves, and the curve graph supports zooming and panning operations. The water temperature curve is represented by a red line, the water level curve is represented by a blue line, and the turbidity curve is represented by a yellow line. The time range of the coordinate axis can be adjusted. The bottom area lists real-time alarm lists, and alarm entries are arranged in reverse chronological order. Unconfirmed alarm entries are highlighted and flickered. The local display module is connected to the industrial control module and the intelligent analysis module. The data refresh mechanism uses a subscription and publication mode. The industrial control module publishes device state change events, and the intelligent analysis module publishes parameter optimization events. The local display module subscribes to these events and updates the interface display. The data export function allows users to save historical data through a USB interface. The data format is selected as the general CSV format, and the file naming rule includes the device number and date range. The export progress bar displays the file generation progress. The touch screen driver supports gesture operations. Single-finger clicking is used for selecting objects, double-finger pinching is used for zooming the curve graph, and long-pressing pops up a context menu for parameter setting and instruction input. The parameter setting interface provides form validation functionality. The value input box limits the value range, and inputs exceeding the range are rejected and an error reason is prompted.

[0055] An adaptive feedback control loop is established between the industrial control module and the intelligent analysis module. The architecture of the adaptive feedback control loop is based on the principle of closed-loop control. The industrial control module sends device running state data packets to the intelligent analysis module in real time. The data packets contain motor current actual values, electric valve opening feedback, and sensor measured values, and the sending frequency is 10 times per second. The intelligent analysis module adjusts the cleaning model parameters according to the received data. The adjustment method uses the recursive least squares method to identify the system model online, and the model parameters are converged to recalculate the optimal set value. The new optimization instruction is issued to the industrial control module through a low-delay communication link. The communication delay compensation module estimates the instruction transmission delay, and the delay estimation value is used to compensate the timestamp of the control instruction, so that the instruction execution time is matched with the system dynamics. The feedback loop uses a PID control algorithm to dynamically correct the control deviation. The proportional coefficient, integral time, and differential time parameters of the PID control algorithm are set by the Ziegler-Nichols method. The control deviation is defined as the difference between the optimal set value and the measured value of the process parameter detection module. The output of the PID control algorithm is used as a feedforward signal to superimpose on the basic control output of the industrial control module. After detecting network timeout, the fault tolerance processing mechanism of communication interruption switches the industrial control module to the locally preset control strategy, maintains the basic operation of the vegetable cleaning machine, and automatically switches back to the optimization control mode after the communication is restored.

[0056] The report generation function of the cloud service module is based on the JasperReports tool, the report template designer defines the layout and style of the report, and the data source is configured as a relational database that stores historical data. Users can specify the time range and report type through the RESTful API interface, and the report generation service generates a PDF format report file asynchronously in the background. The file download link after generation is notified to the user by email. The report content includes device operation statistics, energy consumption analysis and maintenance recommendations, and the chart types include column chart, pie chart and trend line chart. The identity authentication and authorization module controls user access to reports, and the role-based access control model defines different data views for administrators, operators and visitors. The software architecture of the local display module uses the Model-View-Controller design pattern, the model layer encapsulates data access logic, the view layer manages graphical user interface rendering, and the controller layer handles user input and business logic. Software modules communicate through the signal-slot mechanism, and the state change of interface elements emits signals, and slot functions respond to signals to perform corresponding operations. The multi-language support function allows the interface text to be switched to different languages, and the language resource files are stored in separate language packs. The calibration data of the touch screen is stored in non-volatile memory, and the calibration program is started through the hardware test mode, and the user completes the calibration by clicking the target points in turn according to the screen prompts. The performance monitoring of the adaptive feedback control loop records the control quality indicators, including the root mean square value of the set value tracking error, the change amplitude of the control output and the system stability time. Performance monitoring data is uploaded to the cloud service module storage regularly, and long-term trend analysis is used to evaluate the degradation of the cleaning model. When the control quality drops below the threshold, the system prompts the model to update. Maintenance personnel view performance monitoring charts through the local display module or remote client, and the charts show the historical performance and current state of the control loop, assisting in decision-making on whether to reset PID parameters or retrain the cleaning model.

[0057] It should be noted that the relational terms herein such as first and second and the like are used solely to distinguish one from another entity or action, without necessarily requiring or implying that there is any such relationship or order between such entities or actions. Moreover, the terms "comprises", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus.

[0058] Although embodiments of the present application have been shown and described, it is to be understood that various modifications, substitutions, replacements and changes can be made to these embodiments without departing from the principles and spirit of the present application, and the scope of the present application is defined by the appended claims and their equivalents.

Claims

1. A smart vegetable washing machine monitoring system based on the Internet of Things, characterized in that, include: The equipment status monitoring module collects real-time equipment status information of the vegetable washing machine, including power-on status, remote control mode, working status, and fault alarm status. The process parameter detection module monitors water temperature, water level, and turbidity parameters in real time during the cleaning process; The industrial control module, based on equipment status information and process parameters, executes start-stop operations, sequential start-stop control, equipment interlock control, and emergency stop control of the vegetable washing machine. The data integration and transmission module receives real-time data from the equipment status monitoring module and the process parameter detection module, and sends the data to the intelligent analysis module through a wireless communication network; The intelligent analysis module uses the received data to build a dynamic model of the vegetable washing process and generate instructions to optimize the washing parameters. The cloud service module stores historical and real-time data and provides remote access and monitoring functions.

2. The IoT-based intelligent vegetable washing machine monitoring system according to claim 1, characterized in that, The equipment status monitoring module collects equipment status information through multiple sensors integrated on the vegetable washing machine, including using a current sensor to detect the motor's power-on status, a voltage sensor to monitor power supply stability, and a vibration sensor to analyze the equipment's mechanical operating status. The equipment status monitoring module has a built-in signal conversion unit that amplifies and filters the analog signals output by the sensors, converting them into digital signals. The digital signals are analyzed in real time by an embedded microcontroller to extract equipment status feature values, including working cycle, fault frequency, and remote mode identifier. The analyzed data is packaged and transmitted to the industrial control module via a wireless communication protocol.

3. The IoT-based intelligent vegetable washing machine monitoring system according to claim 1, characterized in that, The process parameter detection module uses multi-sensor fusion technology to monitor cleaning process parameters, including real-time water temperature measurement using a platinum resistance temperature sensor, water level detection using an ultrasonic water level sensor, and water turbidity analysis using an optical turbidity sensor. The process parameter detection module is equipped with a data acquisition card to collect sensor data at a fixed sampling frequency and perform analog-to-digital conversion. The converted data is processed by a calibration algorithm to compensate for environmental errors and generate standardized parameter values. The parameter values ​​are sent to the data integration and transmission module via the serial communication interface.

4. The IoT-based intelligent vegetable washing machine monitoring system according to claim 1, characterized in that, The industrial control module implements control logic based on a programmable logic controller (PLC) architecture. This includes writing ladder diagram programs to define equipment start-up and stop sequences, setting timers to control the start-up and stop timing, starting with the equipment at the end of the cleaning line and stopping with the equipment at the beginning. The industrial control module integrates a digital input / output module to receive fault signals from the equipment status monitoring module, triggering interlocking control programs to automatically stop upstream equipment. An emergency stop circuit achieves rapid power-off via a hardware relay. The control module also includes a communication interface to receive optimization instructions from the intelligent analysis module and dynamically adjust control parameters.

5. The IoT-based intelligent vegetable washing machine monitoring system according to claim 1, characterized in that, The data integration and transmission module uses a gateway device to achieve data aggregation and forwarding, including configuring a multi-protocol converter, supporting RS485 and LoRa communication protocols, and parsing data packets from the equipment status monitoring module and the process parameter detection module; The data integration and transmission module runs a data compression algorithm to reduce transmission bandwidth and adds timestamps and data check codes; the compressed data is transmitted to the intelligent analysis module via TCP / IP protocol to ensure data integrity and real-time performance; the gateway device also has a caching mechanism to temporarily store data when the network is interrupted.

6. The IoT-based intelligent vegetable washing machine monitoring system according to claim 1, characterized in that, The intelligent analysis module uses deep learning algorithms to establish a dynamic model of the cleaning process, including collecting historical equipment status and process parameter data as a training set, using convolutional neural networks to extract features, and predicting cleaning efficiency trends. The intelligent analysis module receives streaming data from the data integration and transmission module in real time, and calculates the optimal water temperature, water level and turbidity setpoints through the inference engine; the optimization command generation module adjusts the parameters iteratively using the gradient descent method based on the prediction results, and outputs control commands to the industrial control module.

7. The IoT-based intelligent vegetable washing machine monitoring system according to claim 1, characterized in that, The cloud service module is deployed on a cloud server to realize data storage and remote access, including configuring relational database storage device status records, process parameter history and alarm logs; The cloud service module provides a RESTful API interface, allowing remote users to query real-time data and generate reports; the data synchronization module periodically pulls optimization records from the intelligent analysis module for backup and version management; the cloud service module also integrates identity authentication and encrypted transmission to ensure data security.

8. The IoT-based intelligent vegetable washing machine monitoring system according to claim 1, characterized in that, The system also includes a local display module that visualizes data via an embedded touchscreen, including a graphical user interface for displaying device status diagrams, process parameter curves, and alarm lists; the local display module connects to the industrial control module and the intelligent analysis module to refresh data in real time. The data export function allows users to save historical data via USB interface; the touch screen driver supports gesture operation for parameter setting and command input.

9. The IoT-based intelligent vegetable washing machine monitoring system according to claim 1, characterized in that, An adaptive feedback control loop is established between the industrial control module and the intelligent analysis module. This includes the industrial control module sending the equipment operating status to the intelligent analysis module in real time, the intelligent analysis module adjusting the cleaning model parameters based on the received data, and generating new optimization instructions. The feedback loop uses a PID control algorithm to dynamically correct control deviations and ensure system stability. The communication delay compensation module estimates the transmission delay and synchronizes the control instructions.

10. A monitoring method for an intelligent vegetable washing machine based on the Internet of Things, characterized in that, It includes all modules and method flows of the IoT-based intelligent vegetable washing machine monitoring system as described in any one of claims 1 to 9.

Images