On-line monitoring system and method for sewage treatment, water purification and water turbine power generation

Abstract

The present application discloses a sewage treatment and water purification power generation water turbine online monitoring system, comprising a signal acquisition module, a central control unit module and a warning module; the signal acquisition module is responsible for collecting monitoring data; the signal acquisition module sends the collected data to the Ali cloud platform and stores it in the cloud database; the central control unit module analyzes and processes the monitoring data, and real-time displays the monitoring data obtained by each sensor and the time-frequency domain modal distribution; the Ali cloud platform issues instructions to remotely control the working state of the signal acquisition module and the sampling frequency of the monitoring data; the warning module is triggered by fault information to provide warning information for operation and maintenance personnel. The present application also provides a sewage treatment and water purification power generation water turbine online monitoring method, which solves the problems of difficult water turbine runner vibration signal acquisition and high acquisition cost of the existing sewage treatment and water purification power generation system.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of water conservancy and hydropower, and particularly relates to an online monitoring system for a sewage treatment and water purification hydro turbine, and also provides an online monitoring method for the sewage treatment and water purification hydro turbine. BACKGROUND

[0002] The sewage such as urban production, living water and natural precipitation is collected and purified to reach the standard, and then directly discharged into the nearby river. The discharged reclaimed water has certain potential energy and kinetic energy, and the annual discharge time is as high as 7500 hours or more. Through the water turbine generator, the water energy of the purified water can be successfully recycled, and the online monitoring of the core component of the water turbine is of certain significance. The vibration signal contains important information for mechanical state detection and fault diagnosis, and accurate measurement is helpful for early warning of equipment failure.

[0003] In view of the defects of the traditional online monitoring of the water turbine, the defects are as follows: 1) difficulty in data acquisition, which needs professional engineers to measure; 2) limited data processing capacity, which can only simply analyze part of the data; 3) lack of early warning mechanism, which cannot find potential problems of the water turbine before the failure occurs. In view of the above defects, the purpose of the present application is to combine the traditional vibration sensor acquisition with the Internet of Things transmission, realize the wireless online monitoring of the vibration signal of the water turbine, perform modal processing on the collected data, perform fault diagnosis, thereby solving the problem of difficult data acquisition and transmission, reducing the measurement cost, ensuring the equipment operation, early warning of the equipment failure, thereby avoiding major disasters, and further improving the economic effect. SUMMARY

[0004] The first object of the present application is to provide an online monitoring system for a sewage treatment and water purification hydro turbine, which solves the problem of difficult data acquisition and transmission, reduces the measurement cost, ensures the equipment operation, early warning of the equipment failure, thereby avoids major disasters, and further improves the economic effect.

[0005] The second object of the present application is to provide an online monitoring method for a sewage treatment and water purification hydro turbine.

[0006] The first technical solution adopted by the present application is an online monitoring system for a sewage treatment and water purification hydro turbine, which comprises a signal acquisition module, a central control unit module and an alarm module.

[0007] The signal acquisition module is responsible for collecting monitoring data, and the monitoring data specifically includes the vibration acceleration of the main shaft of the water turbine, the vibration speed of the runner of the water turbine, the pressure pulsation of the tail water pipe wall of the water turbine and the ultrasonic wave of the guide vane crank arm of the water turbine.

[0008] The signal acquisition module sends the collected data to the Aliyun cloud platform and stores the data in the cloud database.

[0009] The central control unit module pulls down the monitoring data received by the Aliyun cloud platform, stores the local data, analyzes and processes the monitoring data, and displays the monitoring data obtained by each sensor and the time-frequency domain mode distribution in real time, and the Aliyun cloud platform issues instructions to remotely control the working state of the signal acquisition module and the sampling frequency of the monitoring data.

[0010] When the central control unit module analyzes and processes the monitoring data, the fault information is extracted, and the alarm module is triggered by the fault information to provide warning information for the operation and maintenance personnel.

[0011] The application is characterized in that,

[0012] The signal acquisition module comprises a signal acquisition and processing submodule, a signal voltage reduction submodule and a sensing and monitoring unit.

[0013] The sensing and monitoring unit comprises a first sensing and monitoring submodule for monitoring the wireless vibration acceleration at the main shaft, a second sensing and monitoring submodule for monitoring the vibration speed at the runner, a third sensing and monitoring submodule for monitoring the pressure pulsation at the tail water pipe wall surface, and a fourth sensing and monitoring submodule for monitoring the ultrasonic wave at the guide vane crank arm.

[0014] The first sensing and monitoring submodule, the second sensing and monitoring submodule, the third sensing and monitoring submodule and the fourth sensing and monitoring submodule transmit the collected monitoring data to the signal acquisition and processing submodule after processing by the signal voltage reduction submodule, and the signal acquisition and processing submodule transmits the monitoring data to the Aliyun cloud platform; the central control unit module issues instructions to remotely control the working state of the first sensing and monitoring submodule, the second sensing and monitoring submodule, the third sensing and monitoring submodule and the fourth sensing and monitoring submodule and the sampling frequency of the corresponding sensing and monitoring module through the Aliyun cloud platform.

[0015] The central control unit module comprises a data analysis and processing submodule, a local data storage function module, a real-time waveform display submodule and a remote control submodule.

[0016] The remote control submodule issues instructions to remotely control the working state of the first sensing and monitoring submodule, the second sensing and monitoring submodule, the third sensing and monitoring submodule and the fourth sensing and monitoring submodule and the sampling frequency of the corresponding sensing and monitoring module through the Aliyun cloud platform.

[0017] The local data storage function module is used for storing the monitoring data uploaded to the Aliyun cloud platform.

[0018] The real-time waveform display submodule displays the monitoring data stored in the Aliyun cloud platform.

[0019] The data analysis processing sub-module performs modal decomposition on the monitoring data transmitted from the Ali cloud platform, determines the frequency distribution of each modal component, classifies each modal component data after decomposition processing, extracts fault information, and transmits the fault information to the warning module for alarm.

[0020] The second technical solution adopted by the application is,

[0021] The specific method for the data analysis processing sub-module to perform modal decomposition on the monitoring data obtained from the Ali cloud platform and determine the frequency distribution of each modal component in step 2 is as follows:

[0022] S1, decompose the monitoring data signal f(t) into K eigenmodes u k , ω k is the center frequency of the eigenmode function, so that the sum of the estimated bandwidth of each eigenmode function is minimized, a constraint variational model is established, and the constraint variational model is specifically:

[0023]

[0024]

[0025] Wherein, formula (1) is a variational model, formula (2) is a constraint condition, t is time,

[0026] δ(t) is the Dirichlet function; f(t) is the monitoring data; K is the number of eigenmodes; u k is the kth modal component; ω k is the center frequency of the kth modal component.

[0027] S2, calculate the center frequency of the eigenmode function:

[0028] Introduce the penalty factor alpha and the time domain Lagrange multiplier lambda(t), and obtain the augmented Lagrange expression as:

[0029]

[0030] Solve the unconstrained variational equation by using the multiplier alternating direction algorithm, and update the iteration u k (ω), ω k (ω), lambda k (ω) continuously;

[0031] S3, use Parseval theorem to solve the eigenmode u k , the center frequency ω k and the Lagrange multiplier lambda(k) in the frequency domain, and the corresponding calculation formula is as follows:

[0032] Eigenmode calculation formula:

[0033]

[0034] Center frequency calculation method:

[0035]

[0036] Lagrange multiplier updating method:

[0037]

[0038] Where, f (ω) is the complex transform of f (t), u k (ω) is the complex transform of u k (t), ω k (ω) is the complex transform of ω k (t), λ k (ω) is the complex transform of λ k (t) ;

[0039] S4, repeat the above formula (4)-(6), respectively, to each road monitoring data signal modal decomposition, each modal center frequency is obtained; Fourier transform is carried out to the obtained each mode, and it is mapped to the frequency domain, and each mode frequency variation trend is obtained.

[0040] The beneficial effects of the application are:

[0041] (1) the method solves the problems of difficult collection of water turbine runner vibration signal and high cost of collection in the existing water treatment and power generation system.

[0042] (2) the method can study the frequency domain variation trend of each mode under different working conditions by processing the monitoring data signal.

[0043] (3) the system reduces the cost of online monitoring of water turbine by using internet of things technology, realizes fast and reliable real-time monitoring of water turbine running state, solves the problems of difficult data acquisition and transmission of rotating equipment in practical engineering, can help staff to master the equipment running state, realize early warning of fault, guide operation personnel to operate and maintain, and provide reference for operation and maintenance personnel. BRIEF DESCRIPTION OF DRAWINGS

[0044] Figure 1 The working flow chart of the central control unit module in the system;

[0045] Figure 2 The VMD algorithm decomposition flow chart in the method;

[0046] Figure 3 The runner vibration acceleration signal time domain graph in the application;

[0047] Figure 4 The vibration acceleration signal frequency domain graph of the runner in the application;

[0048] Figure 5 The VMD decomposition frequency domain of the runner vibration signal when the unit is normal in the application;

[0049] Figure 6 The VMD decomposition time domain of the runner vibration signal when the unit is normal in the application;

[0050] Figure 7 The VMD decomposition time domain graph of the main shaft vibration signal of the unit shaft misalignment in the application;

[0051] Figure 8 The VMD decomposition frequency domain graph of the main shaft vibration signal of the unit shaft misalignment in the application. DETAILED DESCRIPTION

[0052] The application will be described in detail below in combination with the drawings and specific embodiments.

[0053] The application provides a sewage treatment, water purification and power generation water turbine online monitoring system, which comprises a signal acquisition module, a central control unit module and a warning module.

[0054] The signal acquisition module is responsible for acquiring monitoring data, and the monitoring data specifically includes the vibration acceleration of the water turbine main shaft, the vibration speed of the water turbine runner, the pressure pulsation of the water turbine tail water pipe wall surface and the ultrasonic wave of the water turbine guide vane crank arm.

[0055] The signal acquisition module sends the acquired data to the Aliyun cloud platform through wifi and stores the data in the cloud database.

[0056] After networking, the central control unit module pulls down the monitoring data received by the Aliyun cloud platform for local data storage, simultaneously performs data analysis and processing on the monitoring data, and displays the monitoring data obtained by each sensor and the time-frequency domain modal distribution in real time. The Aliyun cloud platform issues instructions to remotely control the working state of the signal acquisition module and the sampling frequency of the monitoring data.

[0057] When the central control unit module finishes analyzing and processing the monitoring data, the fault information is extracted, the warning module is triggered through the fault information, and warning information is provided for the operation and maintenance personnel.

[0058] The signal acquisition module comprises a signal acquisition and processing submodule, a signal step-down submodule and a sensing and monitoring unit.

[0059] The sensing monitoring unit comprises a first sensing monitoring sub-module for monitoring wireless vibration acceleration at the main shaft, a second sensing monitoring sub-module for monitoring vibration speed at the runner, a third sensing monitoring sub-module for monitoring pressure pulsation at the wall surface of the draft tube, and a fourth sensing monitoring sub-module for monitoring ultrasonic waves at the draft tube elbow;

[0060] The first sensing monitoring sub-module, the second sensing monitoring sub-module, the third sensing monitoring sub-module, and the fourth sensing monitoring sub-module respectively transmit the collected monitoring data to the signal acquisition and processing sub-module after processing by the signal step-down sub-module, and the signal acquisition and processing sub-module transmits the monitoring data to the Ali Cloud platform; the central control unit module remotely controls the working state of the first sensing monitoring sub-module, the second sensing monitoring sub-module, the third sensing monitoring sub-module, and the fourth sensing monitoring sub-module and the sampling frequency of the corresponding sensing monitoring module by issuing instructions to the signal acquisition and processing sub-module through the Ali Cloud platform.

[0061] The central control unit module comprises a data analysis and processing sub-module, a local data storage function module, a real-time waveform display sub-module, and a remote control sub-module;

[0062] The remote control sub-module remotely controls the working state (including whether to sleep or intermittently work) of the first sensing monitoring sub-module, the second sensing monitoring sub-module, the third sensing monitoring sub-module, and the fourth sensing monitoring sub-module and the sampling frequency of the corresponding sensing monitoring module by issuing instructions to the signal acquisition and processing sub-module through the Ali Cloud platform;

[0063] The local data storage function module is used for storing the monitoring data uploaded to the Ali Cloud platform;

[0064] The real-time waveform display sub-module displays the monitoring data stored in the Ali Cloud platform;

[0065] The data analysis and processing sub-module performs modal decomposition on the monitoring data transmitted from the Ali Cloud platform, determines the frequency distribution of each modal component, classifies the decomposed modal component data, extracts fault information, and transmits the fault information to the warning module for alarm.

[0066] The application also provides a sewage treatment, water purification, and power generation water turbine online monitoring method, which adopts the sewage treatment, water purification, and power generation water turbine online monitoring system.

[0067] Step 1: The first sensing monitoring sub-module, the second sensing monitoring sub-module, the third sensing monitoring sub-module, and the fourth sensing monitoring sub-module are respectively used to monitor and collect wireless vibration acceleration data at the main shaft, vibration speed data at the runner, pressure pulsation data at the wall surface of the draft tube, and ultrasonic wave data at the draft tube elbow;

[0068] Step 2: If Figure 1As shown, the monitoring data collected in step 1 is transmitted to the signal acquisition and processing submodule after being processed by the signal voltage reduction submodule, and the signal acquisition and processing submodule transmits the monitoring data to the Aliyun cloud platform, which transmits the data to the remote control submodule, the data analysis and processing submodule, and the local data storage function module in the central control unit module; the remote control submodule issues instructions to the signal acquisition and processing submodule through the Aliyun cloud platform to remotely control the working state (including whether to sleep, whether to work intermittently) of the first, second, third, and fourth sensing monitoring submodules and the sampling frequency of the corresponding sensing monitoring modules.

[0069] The data analysis and processing submodule performs modal decomposition on the monitoring data obtained from the Aliyun cloud platform, determines the frequency distribution of each modal component, classifies the decomposed modal component data, extracts fault information, and transmits the fault information to the warning module for alarm; and displays the monitoring data stored in the Aliyun cloud platform through the real-time waveform display submodule.

[0070] In step 2, the data analysis and processing submodule performs modal decomposition on the monitoring data obtained from the Aliyun cloud platform to determine the specific method of determining the frequency distribution of each modal component as follows:

[0071] The decomposition principle is the optimized variational modal decomposition algorithm (VMD), and the detailed decomposition process is shown in Figure 2 As shown:

[0072] S1, decompose the monitoring data signal f(t) into K intrinsic mode functions u k , ω k is the center frequency of the intrinsic mode function (IMF), and each intrinsic mode function has its own frequency bandwidth. In order to minimize the sum of the estimated bandwidth of each intrinsic mode function, a constrained variational model is established, and the constrained variational model is specifically:

[0073]

[0074]

[0075] Wherein, formula (1) is a variational model, formula (2) is a constraint condition, t is time, δ(t) is a Dirac function; f(t) is monitoring data; K is the number of intrinsic modes; u k is the kth modal component; ω k is the center frequency of the kth modal component.

[0076] S2, calculate the intrinsic mode function center frequency:

[0077] The penalty factor α and the time domain Lagrange multiplier λ(t) are introduced, the penalty factor α is used to reduce the Gaussian noise interference and ensure the signal reconstruction accuracy in strong noise, and the Lagrange multiplier λ(t) is used to ensure the strictness of the constraint condition, and the augmented Lagrange expression is obtained as:

[0078]

[0079] The unconstrained variation equation is solved by using the multiplier alternating direction algorithm, and the iteration u k (ω) is continuously updated k (ω) and λ k (ω);

[0080] S3, the Parseval theorem is used to solve the intrinsic mode u k , the center frequency ω k and the Lagrange multiplier λ(k) in the frequency domain, and the corresponding calculation formula is as follows:

[0081] The intrinsic mode calculation formula is:

[0082]

[0083] The center frequency calculation method is:

[0084]

[0085] The Lagrange multiplier updating method is:

[0086]

[0087] Wherein, f(ω) is the complex transform of f(t), u k (ω) is the complex transform of u k (t), ω k (ω) is the complex transform of ω k (t), λ k (ω) is the complex transform of λ k (t);

[0088] In order to further improve the quality of modal decomposition, the difference optimization algorithm is used to optimize the VMD algorithm, and the minimum modal envelope entropy is used as the fitness function, and the number of decomposition layers and the penalty factor are used as the optimization parameters.

[0089] S4, repeat the above formula (4)-(6), respectively, to each monitoring data signal modal decomposition, and obtain the center frequency of each mode; Fourier transform is carried out on the obtained each mode, and it is mapped to the frequency domain, and the frequency change trend of each mode is obtained.

[0090] Embodiment

[0091] The signal acquisition and processing submodule in the signal acquisition module in the sewage treatment, water purification, power generation and water turbine online monitoring system completes the corresponding function through the signal acquisition module chip; the signal voltage reduction submodule completes the corresponding function through the preamplification circuit; the first sensing monitoring submodule completes the corresponding function through the wireless vibration acceleration sensor (model YK-YD500); the second sensing monitoring submodule completes the corresponding function through the vibration speed sensor (YK-YD20); the third sensing monitoring submodule completes the corresponding function through the pressure pulsation sensor (HM90); and the fourth sensing monitoring submodule completes the corresponding function through the ultrasonic sensor (UB1000-18GM55-E4). Specifically, the signal acquisition module chip is ESP32, which is equipped with an Xtensa 32-bit LX7 dual-core processor, and the maximum working frequency can reach 240MHz; the maximum Flash can reach 16MB; the chip includes a complete Wi-Fi subsystem, which complies with the IEEE 802.1b / g / n protocol; the chip has two 13-bit ADC acquisition circuits, and a total of 20 channels can realize multi-channel ADC acquisition; in order to realize the data interaction between the sensor and the chip, a preamplification circuit is designed, the preamplification circuit chip is AD620, the 7th pin of AD620 is connected with the positive electrode of the lithium battery, the 4th pin is connected with the negative electrode of the lithium battery, a 1kΩ resistor R2 is connected in parallel with the 2nd pin and the 3rd pin, a 2kΩ resistor R1 is connected in series with the upper end of R2, the other end of R1 is connected with the output end of the sensor, a 2kΩ resistor R3 is connected in series with the lower end of R2, one end of R3 is connected with the chip GND. A 100MΩ resistor R4 is connected in series between the 1st pin and the 8th pin of AD620, and the 6th pin is connected with the signal acquisition module chip GPIO13.

[0092] The wireless vibration acceleration sensor is installed at the main shaft, the vibration speed sensor is installed at the runner, the pressure pulsation sensor is installed on the wall surface of the draft tube, and the ultrasonic sensor is installed at the guide vane crank arm; the vibration acceleration sensor, the ultrasonic sensor, the pressure pulsation sensor and the vibration sensor respectively transmit the collected data to the signal acquisition module chip through the preamplification circuit, the signal acquisition module chip transmits the monitoring data to the Aliyun cloud platform, and stores them in the cloud database.

[0093] The corresponding functions of the central control unit module are completed by using a PC client with a Windows 10 system, after the computer is networked, the monitoring data received by the Ali cloud platform is pulled down, local data storage is performed, the monitoring data is analyzed and processed, the monitoring data obtained by each sensor and the time-frequency domain mode distribution are displayed in real time, and the signal acquisition module is controlled by the Ali cloud platform to issue instructions to remotely control the working state of the signal acquisition module and the sampling frequency of the monitoring data; wherein, 1) the data processing function corresponds to the VMD grading process in the method of the application, and the data analysis function corresponds to the XGBoost algorithm model in the application; 2) the local data storage function can be used to select the start time and the end time, and the corresponding data can be saved to the local; 3) real-time waveform display, the GUI interface is developed by combining HTML5, JavaScript, CSS3 and ECharts. Figure 3 is a runner vibration acceleration signal time domain graph, and the graph shows the water turbine runner vibration speed curve within 0.4s, the signal changes complexly, and the swing amplitude is large, Figure 4 is a runner vibration acceleration signal frequency domain graph, and the graph shows the vibration acceleration signal frequency domain graph within 0.4s; the graph is a vibration acceleration signal frequency domain graph within 0.4s, the signal frequency component is more, the 500Hz and 3000Hz component amplitudes are the largest, direct analysis and comparison cannot be performed, and the characteristics of each mode need to be further decomposed and extracted; Figure 1 is the wake-up sensor function in the remote control effect.

[0094] A constant voltage source is used to supply power to the PC client, and lithium batteries are used to supply power to each sensor, AD620 and ESP32 chip in the signal acquisition module.

[0095] The alarm module uses a relay to realize the corresponding function, when the signal processing of the central control unit module is completed, if the data waveform is inconsistent, the relay is triggered to alarm, and warning information is provided for the operation and maintenance personnel.

[0096] In use, first, the sensors are arranged at different parts of the water turbine, the wireless vibration acceleration sensor is installed on the main shaft, the vibration speed sensor is installed on the runner, the pressure pulsation sensor is installed on the wall surface of the draft tube, and the ultrasonic sensor is installed on the guide vane arm; then, the system is started, the PC client computer is networked, the signal acquisition module chip starts networking, the network connection is completed, the PC client issues a wake-up command for each sensor, the corresponding signal acquisition system is woken up, the ADC starts signal acquisition, and data storage is completed. When the stop connection command is issued, each signal acquisition system enters a sleep state, and the working process is as shown in Figure 1The PC client computer carries out modal decomposition on each monitoring data signal, determines the frequency distribution of each modal component, and obtains each modal center frequency; the obtained each mode is subjected to Fourier transform, is mapped to the frequency domain, and the frequency variation trend of each mode is studied. Figure 5 is a VMD decomposition frequency domain graph of the runner vibration signal when the unit is normal, and the figure is a VMD decomposition frequency domain graph of the water turbine in normal operation, the signal contains four main modes, and the distribution span of each mode is large. Figure 6 is a VMD decomposition time-frequency graph of the runner vibration signal when the unit is normal, and the figure is a VMD decomposition time domain graph of the water turbine in normal operation, the signal contains four main modes, and the distribution span of each mode is large; Figure 7 is a VMD decomposition time domain graph of the main shaft vibration signal of the unit shaft misalignment; Figure 8 is a VMD decomposition frequency domain graph of the main shaft vibration signal of the unit shaft misalignment; when the shaft is misaligned, the number of signal modes is reduced, and the frequency band of each mode is shifted, and according to the change, the running condition of the unit runner can be distinguished. By observing the frequency change of the signal, the unit shaft misalignment fault can be found. In order to further improve the accuracy and intelligent judgment of the online detection system, the XGBoost multi-classification algorithm of the deep learning model is introduced into the present application, and a fault evaluation model is established to provide early warning for the fault. The fault evaluation model takes the integrated deep model as the basic model, optimizes the combination of the model, and adaptively constructs the fault evaluation model to pre-judge the fault for the reference of the operation and maintenance personnel. The problem of difficult data acquisition and transmission is solved, the measurement cost is reduced, the equipment operation is guaranteed, the early warning of the equipment fault is realized, the major disaster is avoided, and the economic effect is improved.

Claims

1. An online monitoring system for sewage treatment, water purification, and hydroelectric turbine, characterized in that, The signal acquisition module, the central control unit module and the warning module are included. The signal acquisition module is responsible for collecting monitoring data, and the monitoring data is specifically vibration acceleration of a water turbine main shaft, vibration speed of a water turbine runner, pressure pulsation of a water turbine draft tube wall surface and ultrasonic waves of a water turbine guide vane crank arm. The signal acquisition module sends the collected data to the Ali Cloud platform and stores it in the cloud database. The central control unit module pulls down the monitoring data received by the Ali Cloud platform, stores the data locally, analyzes and processes the monitoring data, and displays the monitoring data obtained by each sensor and the time-frequency domain modal distribution in real time. When the central control unit module finishes analyzing and processing the monitoring data, the fault information is extracted, and the warning module is triggered by the fault information to provide warning information for operation and maintenance personnel. The signal acquisition module includes a signal acquisition and processing sub-module, a signal step-down sub-module and a sensing and monitoring unit. The sensing and monitoring unit includes a first sensing and monitoring sub-module for monitoring wireless vibration acceleration at the main shaft, a second sensing and monitoring sub-module for monitoring vibration speed at the runner, a third sensing and monitoring sub-module for monitoring pressure pulsation at the draft tube wall surface, and a fourth sensing and monitoring sub-module for monitoring ultrasonic waves at the guide vane crank arm. The first, second, third and fourth sensing and monitoring sub-modules respectively transmit the collected monitoring data to the signal acquisition and processing sub-module after processing by the signal step-down sub-module, and the signal acquisition and processing sub-module transmits the monitoring data to the Ali Cloud platform. The central control unit module remotely controls the working state of the first, second, third and fourth sensing and monitoring sub-modules and the sampling frequency of the corresponding sensing and monitoring modules by issuing instructions to the signal acquisition and processing sub-module through the Ali Cloud platform. The central control unit module includes a data analysis and processing sub-module, a local data storage function module, a real-time waveform display sub-module and a remote control sub-module. The remote control sub-module remotely controls the working state of the first, second, third and fourth sensing and monitoring sub-modules and the sampling frequency of the corresponding sensing and monitoring modules by issuing instructions to the signal acquisition and processing sub-module through the Ali Cloud platform. The local data storage function module is used to store the monitoring data uploaded to the Ali Cloud platform. The real-time waveform display sub-module displays the monitoring data stored in the Ali Cloud platform.

2. The method for on-line monitoring of sewage treatment, water purification, and hydroelectric turbine, characterized in that, The data analysis and processing sub-module performs modal decomposition on the monitoring data transmitted from the Ali Cloud platform, determines the frequency distribution of each modal component, classifies the decomposed and processed modal component data, extracts fault information, and transmits the fault information to the warning module for alarm. The sewage treatment, water purification and power generation water turbine online monitoring system of claim 1 comprises the following specific steps: Step 1, the first sensing monitoring sub-module, the second sensing monitoring sub-module, the third sensing monitoring sub-module and the fourth sensing monitoring sub-module are used for monitoring and collecting wireless vibration acceleration data at the main shaft, vibration speed data at the runner, pressure fluctuation data at the draft tube wall surface and ultrasonic wave data at the draft tube arm respectively; Step 2, the monitoring data collected in step 1 is transmitted to the signal acquisition and processing sub-module after being processed by the signal step-down sub-module, and the signal acquisition and processing sub-module transmits the monitoring data to the Ali cloud platform, and the Ali cloud platform transmits the data to the remote control sub-module, the data analysis and processing sub-module and the local data storage function module in the central control unit module; the remote control sub-module controls the working state of the first sensing monitoring sub-module, the second sensing monitoring sub-module, the third sensing monitoring sub-module and the fourth sensing monitoring sub-module and the sampling frequency of the corresponding sensing monitoring module through the Ali cloud platform; The data analysis and processing sub-module performs modal decomposition on the monitoring data obtained from the Ali cloud platform, determines the frequency distribution of each modal component, classifies each modal component data after decomposition processing, extracts fault information, and transmits the fault information to the warning module for alarm; And the monitoring data stored in the Ali cloud platform is displayed through the real-time waveform display sub-module; In step 2, the specific method of the data analysis and processing sub-module for modal decomposition on the monitoring data obtained from the Ali cloud platform to determine the frequency distribution of each modal component is as follows: S1, decompose the monitoring data signal f(t) into K eigenmode functions u k , ω k is the center frequency of the eigenmode function, the sum of the estimated bandwidth of each eigenmode function is minimized, a constrained variational model is established, and the constrained variational model is specifically: wherein formula (1) is a variational model, formula (2) is a constraint condition, t is time, δ(t) is a Dirichlet function; f(t) is monitoring data; K is the number of eigenmodes; u k is the kth modal component; ω k is the central frequency of the kth modal component; S2, calculate the central frequency of the intrinsic modal function: Introduce the penalty factor alpha and the time domain Lagrange multiplier lambda(t), and get the augmented Lagrange expression as: The unconstrained variational equation is solved by using the multiplier alternating direction algorithm, and the iteration u k (ω), ω k (ω), λ k (ω); S3. Use Parseval's theorem to analyze the eigenmode u in the frequency domain. k , center frequency ω k And Lagrange multiplier λ(k) are solved, and the corresponding calculation formula is as follows: Intrinsic modal calculation formula: Central frequency calculation method: Lagrange multiplier updating method: where f(ω) is the complex transform of f(t), u k (ω) is the complex transform of u k (t), ω k (ω) is the complex transform of ω k (t), λ k (ω) is the complex transform of λ k (t). S4, repeat the above formulas (4)-(6), respectively, to modal decomposition of each monitoring data signal, to obtain the central frequency of each mode; Fourier transform is performed on the obtained each mode to map it to the frequency domain, and the frequency variation trend of each mode is obtained.

Images