A diesel generator set remote monitoring system

Abstract

This invention discloses a remote monitoring system for diesel generator sets, belonging to the field of fault monitoring technology. It solves the problem of difficulty in effectively diagnosing winding overheating and insulation degradation caused by harmonics in backup diesel generator sets in data centers. This leads to a vicious cycle of harmonic heating and accelerated insulation aging, potentially resulting in sudden breakdown and burnout of the generator windings without any clear warning. The system includes a dielectric constant sensor network, a flexible capacitive sensor array, an insulation state sensing module, a harmonic stress field reconstruction module, an adaptive signal injection module, a central processing module, and a closed-loop execution module. Through multi-dimensional direct sensing and dynamic model calculation, this invention transforms the implicit vicious cycle of winding overheating and insulation aging caused by harmonics into an explicit management process that can be continuously observed, quantitatively assessed, and proactively intervened in, thereby providing clear early warnings and mitigation measures before insulation breakdown occurs.

Description

Technical Field

[0001] This invention relates to the field of fault monitoring technology, and in particular to a remote monitoring system for diesel generator sets. Background Technology

[0002] Diesel generator sets are complex mechatronic systems integrating mechanical and electrical components. Their function is to convert the chemical energy of diesel fuel into electrical energy, providing power for various scenarios. They mainly consist of a diesel engine, a synchronous generator, a control system, auxiliary systems, and a power distribution system. The diesel engine outputs mechanical torque as the power source, while the synchronous generator achieves energy conversion through electromagnetic induction between the stator and rotor. Auxiliary systems (fuel, lubrication, cooling, etc.) ensure stable operation of the unit. The electrical system encompasses core circuits such as main power generation, excitation control, and monitoring and protection. It uses an automatic voltage regulator and controller as its main structure to precisely regulate the output voltage and frequency, monitor parameters such as oil pressure and water temperature in real time, and implement fault protection. It can operate independently or in parallel with multiple units, making it a key device for emergency power supply and scenarios without mains power.

[0003] Backup diesel generator sets in data centers serve as critical emergency power supplies supporting cloud computing centers or large enterprise server rooms. While typically in standby mode, they are vulnerable to power outages when mains power is interrupted. Facing power surges of hundreds of kilowatts to megawatts from nonlinear loads such as server power supplies and inverters, these generators experience high-order harmonics. These harmonic currents cause slight voltage waveform distortion and generate additional resistance and eddy current losses in the generator stator windings, leading to imperceptible localized overheating. Since insulation thermal aging requires offline testing for diagnosis, the degradation process is prolonged, accelerating thermal degradation of the insulation material and reducing its dielectric strength. Insulation aging also increases winding resistance unevenness and sensitivity to harmonics, making it prone to partial discharge. Ultimately, during normal mains power switching or load surges, the diesel generator set may experience complete winding insulation breakdown, resulting in inter-turn or ground short circuits, melting copper wires, and unit burnout, causing a catastrophic power outage.

[0004] Therefore, a remote monitoring system for diesel generator sets is proposed to solve or alleviate the above problems. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of existing technologies by proposing a remote monitoring system for diesel generator sets.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: A remote monitoring system for a diesel generator set includes a dielectric constant sensor network, a flexible capacitor sensor array, an insulation state sensing module, a harmonic stress field reconstruction module, an adaptive signal injection module, a central processing module, and a closed-loop execution module. The dielectric constant sensor network and flexible capacitance sensor array are installed within the diesel generator set. The dielectric constant sensor network measures the capacitance changes of the sensing units embedded in the diesel generator set to reflect the aging state of the molecular structure of the insulating material. The flexible capacitance sensor array monitors the physical state of the stator winding by sensing the capacitance changes caused by minute deformations on the surface of the stator winding due to thermal stress or mechanical vibration. The insulation state sensing module is connected to the dielectric constant sensor network and the flexible capacitance sensor array, and is also connected to the central processing module. The voltage and current input channels of the harmonic stress field reconstruction module are respectively connected to the secondary side of the voltage transformer and the current sensor in the diesel generator set, and its data output terminal is connected to the central processing module. Its synchronous clock input is connected to the system global clock source. The harmonic stress field reconstruction module calculates the harmonic loss density distribution. The injection signal output terminal of the adaptive signal injection module is connected to the output busbar of the diesel generator set. Its response signal input terminal is coupled to the sensor signal link of the harmonic stress field reconstruction module. Its control and data interface is connected to the central processing module. The central processing module performs dynamic reconstruction of the thermal stress field based on the harmonic loss density distribution, estimates the dynamic state of insulation aging by combining data from the insulation state sensing module, and performs adaptive active detection and online calibration of model parameters in conjunction with the adaptive signal injection module. It also performs comprehensive health assessment and dynamic prediction of remaining life. The central processing module is connected to the closed-loop execution module to send control commands. The control signal output terminal of the closed-loop execution module is connected to the active filter and control cabinet of the diesel generator set to receive and execute commands from the central processing module.

[0007] Preferably, the dielectric constant sensor network includes multiple flexible sensing units and a multiplexed data bus. Each flexible sensing unit includes a flexible substrate, interdigitated electrodes disposed on the flexible substrate, and an insulating protective layer covering the interdigitated electrodes. The leads of the two interdigitated electrodes of each flexible sensing unit are respectively connected to the signal line and the common reference ground line in the multiplexed data bus. The terminal of the multiplexed data bus is connected to the measurement input terminal of the capacitance-to-digital converter in the insulation state sensing module.

[0008] Preferably, the flexible capacitive sensor array includes multiple capacitive sensing units, a flexible flat cable, and an integrated connector. Each capacitive sensing unit includes a flexible substrate, parallel plate electrodes disposed on the surface of the flexible substrate, and an elastomer coating layer encapsulating the parallel plate electrodes. The two parallel plate electrodes of each capacitive sensing unit are connected to the integrated connector via the flexible flat cable. The integrated connector is connected to the analog input channel of the capacitance-to-digital converter within the insulation state sensing module.

[0009] Preferably, the insulation state sensing module includes a microcontroller, a capacitor-to-digital converter, a gas sensing signal conditioning circuit, a multi-channel analog-to-digital converter, and a first board-to-board connector. The digital interface of the capacitor-to-digital converter is connected to the microcontroller through a first digital isolator. The output of the gas sensing signal conditioning circuit is connected to the analog input channel of the multi-channel analog-to-digital converter. The digital output interface of the multi-channel analog-to-digital converter is connected to the serial peripheral interface of the microcontroller. The transmit data input and receive data output of the first communication interface chip of the multi-channel analog-to-digital converter are respectively connected to the transmit and receive ends of the microcontroller's universal asynchronous transceiver through a second digital isolator. The microcontroller's universal input / output group, serial peripheral interface, and differential output of the first communication interface chip are all connected to the corresponding communication ends of the first board-to-board connector. The first board-to-board connector is connected to the central processing module.

[0010] Preferably, the harmonic stress field reconstruction module includes a synchronous sampling analog-to-digital converter, a first signal conditioning circuit, a field-programmable gate array (FPGA), a phase-locked loop (PLL) circuit, an external signal connector, and a second board-to-board connector. The voltage input terminal and current input terminal of the external signal connector are respectively connected to the corresponding analog input channel of the synchronous sampling FPGA through the first signal conditioning circuit. The reference clock input terminal of the PLL circuit is connected to the system global clock source, and its output clock terminal is connected to the external sampling clock input terminal of the synchronous sampling FPGA and the global clock input terminal of the FPGA. The digital data output terminal of the synchronous sampling FPGA is connected to the high-speed data input interface of the FPGA. The data bus interface, configuration interface, and general input / output terminal group of the FPGA are all connected to the second board-to-board connector, which is connected to the central processing module.

[0011] Preferably, the adaptive signal injection module includes a direct digital frequency synthesizer, a digital-to-analog converter, a programmable gain amplifier, a power amplifier, an isolation transformer, a high-speed analog-to-digital converter, a second signal conditioning circuit, a directional coupler, a current limiting protection circuit, a transient voltage suppressor, and a third board-to-board connector; the parallel data output port of the direct digital frequency synthesizer is connected to the parallel data input port of the digital-to-analog converter; the analog voltage output port of the digital-to-analog converter is connected in series with the input terminal of the programmable gain amplifier, the input terminal of the power amplifier, and the primary winding of the isolation transformer; the secondary winding of the isolation transformer... After being connected to the current limiting protection circuit and the transient voltage suppressor, it is connected to the generator output bus through a dedicated interface on the front panel of the module for injecting test signals; the coupling end of the directional coupler is connected in series in the secondary signal link of the generator current transformer, and its output end is connected in sequence to the input end of the second signal conditioning circuit and the analog input channel of the high-speed analog-to-digital converter for acquiring the injected response signal; the reference clock input end, configuration bus interface of the direct digital frequency synthesizer, and data output interface of the high-speed analog-to-digital converter are all connected to the third board-to-board connector, which is connected to the central processing module.

[0012] Preferably, the central processing module includes a multi-core processor, a field-programmable gate array (FPGA), a time-sensitive network (TSN) Ethernet switch chip, a non-volatile memory (NSM), a watchdog circuit, and a fourth board-to-board connector. The multi-core processor is connected to the FPGA and the TSN via a high-speed serial computer expansion bus standard channel. Multiple general-purpose input / output (GPIO) groups and a high-speed serial interface of the FPGA are connected to the fourth board-to-board connector. The NSM is connected to the multi-core processor via a serial peripheral interface or a parallel bus. The reset signal output of the watchdog circuit is connected to the system reset terminal of the multi-core processor. The fourth board-to-board connector is connected to the TSN and the FPGA's GPIO groups.

[0013] Preferably, the closed-loop execution module includes an isolated digital output chip, an isolated analog output chip, an optocoupler isolator, a power driver, a digital signal isolator, a communication protocol conversion chip, and a fifth board-to-board connector. The data input port of the isolated digital output chip is connected to the data bus terminal of the fifth board-to-board connector through the digital signal isolator. The digital control interface of the isolated analog output chip is connected to another output channel of the digital signal isolator. Each output channel of the isolated digital output chip is connected to the first set of terminals of the terminal block in sequence through the optocoupler isolator and the power driver. The current output port of the isolated analog output chip is connected to the second set of terminals of the terminal block. The terminal block is used to connect the control input terminal of the active filter and the remote control terminal of the control cabinet in the generator set. One end of the communication protocol conversion chip is connected to the auxiliary communication terminal of the fifth board-to-board connector through a serial bus. The fifth board-to-board connector is connected to the central processing module.

[0014] The present invention has the following beneficial effects: This invention transforms the hidden vicious cycle of winding overheating and insulation aging caused by harmonics into an explicit management process that can be continuously observed, quantitatively evaluated, and proactively intervened in through multi-dimensional direct sensing and dynamic model calculation, thereby providing clear early warning and mitigation measures before insulation breakdown occurs. Attached Figure Description

[0015] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a structural block diagram of the present invention.

[0017] In the diagram: 1. Dielectric constant sensor network; 2. Flexible capacitive sensor array; 3. Insulation state sensing module; 4. Central processing module; 5. Harmonic stress field reconstruction module; 6. Adaptive signal injection module; 7. Closed-loop execution module. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0019] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0020] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0021] In the description of this invention, it should be understood that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship commonly used when the product of this invention is in use, or the orientation or positional relationship commonly understood by those skilled in the art. They are only used to facilitate the description of this invention and to simplify the description, and are not intended to indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0022] Furthermore, the terms "first," "second," and "third" are used only to distinguish descriptions and should not be interpreted as indicating or implying relative importance.

[0023] In the description of this invention, it should also be noted that, unless otherwise explicitly specified and limited, the terms "set," "install," "connect," and "link" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0024] A remote monitoring system for diesel generator sets, such as Figure 1 As shown, it includes a dielectric constant sensor network 1, a flexible capacitive sensor array 2, an insulation state sensing module 3, a harmonic stress field reconstruction module 5, an adaptive signal injection module 6, a central processing module 4, and a closed-loop execution module 7. Dielectric constant sensor network 1 and flexible capacitance sensor array 2 are installed within the diesel generator set. Dielectric constant sensor network 1 measures the capacitance changes of the sensing units embedded in the diesel generator set to reflect the aging state of the molecular structure of the insulating material. Flexible capacitance sensor array 2 monitors the physical state of the stator winding by sensing the capacitance changes caused by minute deformations on the surface of the stator winding due to thermal stress or mechanical vibration. Insulation state sensing module 3 is connected to dielectric constant sensor network 1 and flexible capacitance sensor array 2, and is also connected to central processing module 4. The voltage and current input channels of harmonic stress field reconstruction module 5 are connected to the secondary side of the voltage transformer and the current sensor in the diesel generator set, respectively, and its data output is connected to central processing module 4. Its synchronous clock input is connected to the system global clock source. The harmonic stress field reconstruction module 5 calculates the harmonic loss density distribution. The injection signal output terminal of the adaptive signal injection module 6 is used to connect to the output busbar of the diesel generator set. Its response signal input terminal is coupled to the sensor signal link of the harmonic stress field reconstruction module 5. Its control and data interface is connected to the central processing module 4. The central processing module 4 performs dynamic reconstruction of the thermal stress field based on the harmonic loss density distribution, estimates the dynamic state of insulation aging by combining the data of the insulation state sensing module 3, and performs adaptive active detection and online calibration of model parameters in conjunction with the adaptive signal injection module 6. It also performs comprehensive health assessment and dynamic prediction of remaining life. The central processing module 4 is connected to the closed-loop execution module 7 to send control commands. The control signal output terminal of the closed-loop execution module 7 is connected to the active filter and control cabinet of the diesel generator set to receive and execute the commands from the central processing module 4.

[0025] The dielectric constant sensor network 1 includes multiple flexible sensing units and a multiplexed data bus. Each flexible sensing unit includes a flexible substrate, interdigitated electrodes disposed on the flexible substrate, and an insulating protective layer covering the interdigitated electrodes. The leads of the two interdigitated electrodes of each flexible sensing unit are respectively connected to the signal line and the common reference ground line in the multiplexed data bus. The terminal of the multiplexed data bus is connected to the measurement input terminal of the capacitance digital converter in the insulation state sensing module 3.

[0026] The flexible capacitive sensor array 2 includes multiple capacitive sensing units, a flexible flat cable, and an integrated connector. Each capacitive sensing unit includes a flexible substrate, parallel plate electrodes disposed on the surface of the flexible substrate, and an elastomer coating layer encapsulating the parallel plate electrodes. The two parallel plate electrodes of each capacitive sensing unit are connected to the integrated connector via the flexible flat cable. The integrated connector is connected to the analog input channel of the capacitance-to-digital converter in the insulation state sensing module 3.

[0027] The insulation state sensing module 3 includes a microcontroller STM32G474RET6, a capacitor-to-digital converter AD7747, a gas sensing signal conditioning circuit, a multi-channel analog-to-digital converter ADS131M08, and a first board-to-board connector. The digital interface of the capacitor-to-digital converter is connected to the microcontroller through a first digital isolator ISO7741. The output of the gas sensing signal conditioning circuit is connected to the analog input channel of the multi-channel analog-to-digital converter. The digital output interface of the multi-channel analog-to-digital converter is connected to the serial peripheral interface of the microcontroller. The transmit data input and receive data output of the first communication interface chip of the multi-channel analog-to-digital converter are connected to the transmit and receive ends of the microcontroller's universal asynchronous transceiver through a second digital isolator ISO7741, respectively. The microcontroller's universal input / output group, serial peripheral interface, and differential output of the first communication interface chip are all connected to the corresponding communication ends of the first board-to-board connector. The first board-to-board connector is connected to the central processing module 4.

[0028] The gas sensor signal conditioning circuit includes a gas sensor interface, a constant current source, an instrumentation amplifier INA333, a low-pass filter, and a voltage follower. The positive and negative input terminals of the gas sensor interface are connected to the corresponding terminals of the gas sensor in the diesel generator set, respectively. The positive input terminal is connected to the output terminal of the constant current source. The output terminal of the gas sensor interface is connected to the positive and negative differential input terminals of the instrumentation amplifier. The output terminal of the instrumentation amplifier is connected in series with the low-pass filter and the voltage follower. The output terminal of the voltage follower is connected to a designated analog input channel of the multi-channel analog-to-digital converter. The enable terminal of the constant current source, the reference voltage terminal of the instrumentation amplifier, and the cutoff frequency control terminal of the low-pass filter are respectively connected to specific general-purpose input and output terminals of the microcontroller.

[0029] The harmonic stress field reconstruction module 5 includes a synchronous sampling analog-to-digital converter (ADC) AD7779, a first signal conditioning circuit, a field-programmable gate array (FPGA), a phase-locked loop (PLL) circuit ADF4002, an external signal connector, and a second board-to-board connector. The voltage and current input terminals of the external signal connector are connected to the corresponding analog input channels of the synchronous sampling ADC through the first signal conditioning circuit. The reference clock input terminal of the PLL circuit is connected to the system global clock source, and its output clock terminal is connected to the external sampling clock input terminal of the synchronous sampling ADC and the global clock input terminal of the FPGA. The digital data output terminal of the synchronous sampling ADC is connected to the high-speed data input interface of the FPGA. The data bus interface, configuration interface, and general-purpose input / output terminal group of the FPGA are all connected to the second board-to-board connector. The second board-to-board connector is connected to the central processing module 4.

[0030] The first signal conditioning circuit includes a precision resistor divider network, a current sensing interface, a multi-stage active low-pass filter, and a gain programmable instrumentation amplifier (PGA281). The voltage input terminal of the external signal connector is connected to the input terminal of the multi-stage active low-pass filter through the precision resistor divider network. The input terminal of the current sensing interface is connected to the current input terminal of the external signal connector, and its output terminal is connected to the differential input terminal of the gain programmable instrumentation amplifier. The output terminals of the multi-stage active low-pass filter and the gain programmable instrumentation amplifier are respectively connected to different analog input channels of the synchronous sampling analog-to-digital converter. The gain control terminal of the gain programmable instrumentation amplifier and the cutoff frequency control terminal of the multi-stage active low-pass filter are respectively connected to specific general-purpose input / output terminals of the field-programmable gate array (FPGA).

[0031] The adaptive signal injection module 6 includes a direct digital frequency synthesizer AD9914, a digital-to-analog converter AD5791, a programmable gain amplifier PGA281, a power amplifier, an isolation transformer, a high-speed analog-to-digital converter AD9269, a second signal conditioning circuit, a directional coupler, a current limiting protection circuit, a transient voltage suppressor, and a third board-to-board connector; the parallel data output port of the direct digital frequency synthesizer is connected to the parallel data input port of the digital-to-analog converter; the analog voltage output port of the digital-to-analog converter is connected in series with the input terminal of the programmable gain amplifier, the input terminal of the power amplifier, and the primary winding of the isolation transformer; After the secondary winding of the transformer is connected to the current limiting protection circuit and transient voltage suppressor, it is connected to the generator output bus through a dedicated interface on the front panel of the module for injecting test signals. The coupling end of the directional coupler is connected in series in the secondary signal link of the generator current transformer, and its output end is connected in sequence to the input end of the second signal conditioning circuit and the analog input channel of the high-speed analog-to-digital converter for acquiring the injected response signal. The reference clock input end, configuration bus interface and data output interface of the direct digital frequency synthesizer are all connected to the third board-to-board connector, which is connected to the central processing module 4.

[0032] The second signal conditioning circuit includes a high-pass filter, a programmable gain amplifier ADA4898-2, an anti-aliasing bandpass filter, and a DC bias adjustment circuit. The input of the high-pass filter is connected to the signal output of the directional coupler, and the output of the high-pass filter is connected to the differential signal input of the programmable gain amplifier. The output of the programmable gain amplifier is connected in sequence to the input of the anti-aliasing bandpass filter and the input of the DC bias adjustment circuit. The output of the DC bias adjustment circuit is connected to a designated analog input channel of the high-speed analog-to-digital converter. The gain control terminal of the programmable gain amplifier, the cutoff frequency control terminal of the anti-aliasing bandpass filter, and the bias voltage control terminal of the DC bias adjustment circuit are respectively connected to the control bus terminal of the third board-to-board connector.

[0033] The current limiting protection circuit includes a fast-acting fuse, a Hall effect current sensor, a high-speed voltage comparator, a gate driver, a solid-state relay, and a fault status latch. The fast-acting fuse is connected in series in the output circuit of the secondary winding of the isolation transformer. The measuring aperture of the Hall effect current sensor is sleeved on the current conductor after the fast-acting fuse, and its output voltage terminal is connected to the non-inverting input terminal of the high-speed voltage comparator. The inverting input terminal of the high-speed voltage comparator is connected to a preset current threshold reference voltage introduced by the third board-to-board connector. The output terminal of the high-speed voltage comparator is connected to the input terminal of the gate driver circuit. The output terminal of the gate driver circuit is connected to the control terminal of the solid-state relay. The switching contacts of the solid-state relay are connected in series in the main output circuit of the secondary winding of the isolation transformer. The output terminal of the high-speed voltage comparator is connected to the set terminal of the fault status latch. The data output terminal of the fault status latch is connected to the status feedback terminal of the third board-to-board connector.

[0034] Central Processing Module 4 includes a multi-core processor, a field-programmable gate array (FPGA), a time-sensitive network (TSN) Ethernet switch chip, non-volatile memory, a watchdog circuit, and a fourth board-to-board connector. The multi-core processor is connected to the FPGA and the TSN Ethernet switch chip via a high-speed serial computer expansion bus standard channel. Multiple general-purpose input / output (GPIO) groups and high-speed serial interfaces of the FPGA are connected to the fourth board-to-board connector. The non-volatile memory is connected to the multi-core processor via a serial peripheral interface or a parallel bus. The reset signal output of the watchdog circuit is connected to the system reset terminal of the multi-core processor. The fourth board-to-board connector is connected to the TSN Ethernet switch chip and the FPGA's GPIO groups.

[0035] The closed-loop execution module 7 includes an isolated digital output chip ISO7241C, an isolated analog output chip AD5420, an optocoupler, a power driver, a digital signal isolator ADuM1402, a communication protocol conversion chip MAX3485, and a fifth board-to-board connector. The data input port of the isolated digital output chip is connected to the data bus of the fifth board-to-board connector through the digital signal isolator. The digital control interface of the isolated analog output chip is connected to another output channel of the digital signal isolator. Each output channel of the isolated digital output chip is connected to the first set of terminals of the terminal block through the optocoupler and the power driver in sequence. The current output port of the isolated analog output chip is connected to the second set of terminals of the terminal block. The terminal block is used to connect the control input terminal of the active filter and the remote control terminal of the control cabinet in the generator set. One end of the communication protocol conversion chip is connected to the auxiliary communication terminal of the fifth board-to-board connector through a serial bus. The fifth board-to-board connector is connected to the central processing module 4.

[0036] Furthermore, when the remote monitoring system for this diesel generator set is in operation, it performs the following steps: S1, Insulation state sensing module 3 and harmonic stress field reconstruction module 5 perform multi-source data synchronization and time-scale unified preprocessing; The central processing module 4 sends synchronous acquisition commands to the insulation state sensing module 3, the harmonic stress field reconstruction module 5, and the adaptive signal injection module 6. Each module collects dielectric constant, characteristic gas concentration, voltage and current waveforms, and response signal data based on a unified system clock signal. After receiving data from each module, the central processing module 4 compensates and aligns the timestamps of all data points according to the preset fixed transmission delay parameters of each channel, and generates multimodal time series data with a unified absolute time base. S2, Harmonic stress field reconstruction module 5 performs harmonic loss density distribution calculation; The harmonic stress field reconstruction module 5 performs windowed interpolation fast Fourier transform analysis on the synchronized current waveform data to decompose and obtain the amplitude and phase complex vectors of each integer harmonic current. Based on the DC resistance, geometric dimensions and material properties of the winding conductor, the AC resistance coefficient corresponding to each harmonic is calculated. This coefficient increases with the harmonic order in a power function relationship. Based on the spatial distribution map of the generator winding, the harmonic loss is obtained by multiplying each harmonic current by the corresponding AC resistance, and the loss density is distributed according to the spatial location to form a three-dimensional harmonic loss density distribution map. More specifically, First, a windowed interpolation Fast Fourier Transform (FFT) analysis method is applied to the three-phase current waveforms to reduce spectral analysis errors and improve frequency resolution, thereby obtaining each harmonic, usually represented by the h-th harmonic, in complex form, including amplitude and phase angle. Next, the AC resistance of the windings under different harmonic frequencies is calculated. This resistance value is equal to the DC resistance of the windings multiplied by a coefficient, which consists of three parts: a digit (1); a skin effect coefficient multiplied by the square of the harmonic order; and a proximity effect coefficient multiplied by the square of the harmonic order. The DC resistance is derived from the generator set's design specifications or actual measurements. The skin effect coefficient is related to the conductor diameter, operating frequency, and material conductivity, and can be calculated using specific formulas or manufacturer data. The proximity effect coefficient is related to the winding arrangement and insulation thickness, and is derived from electromagnetic simulation or empirical formulas. Finally, spatial loss density modeling is performed. The loss density per unit volume of each discrete winding unit is equal to one-third of the unit volume multiplied by the sum of the squares of the amplitudes of all harmonic currents flowing through the unit and their corresponding AC resistances, plus the eddy current loss of the unit. The unit volume data comes from the generator set. The harmonic current flowing through a specific unit needs to be allocated from the bus current through the winding distribution model. Eddy current loss is caused by the changing magnetic field in the core and windings, and the harmonic loss density distribution map can be pre-calculated through electromagnetic field finite element analysis.

[0037] S3, Central Processing Module 4 performs dynamic reconstruction of the thermal-stress field based on harmonic loss density distribution; The central processing module 4 inputs the harmonic loss density distribution map as a heat source into its built-in three-dimensional thermal network digital model of the generator set; The thermal network digital model consists of nodes, thermal resistance, and heat capacity. By solving a set of differential equations with node temperature as the variable, the transient three-dimensional temperature field inside the generator set under the combined action of harmonic heat source, cooling wind, and ambient temperature is calculated. The highest temperature value and its spatial coordinates are extracted from the three-dimensional temperature field and used as the hotspot temperature and hotspot location. More specifically, Solve for the time-varying three-dimensional temperature field formed inside the generator set by the losses calculated in step two. The method employed is to establish and solve a three-dimensional thermal network digital model of the generator set. This model is expressed as follows: the node thermal capacity matrix multiplied by the derivative of the node temperature with respect to time is equal to the node thermal conduction matrix multiplied by the ambient temperature vector minus the node temperature vector, plus the heat source distribution matrix multiplied by the volume loss density vector. The node thermal capacity matrix is ​​a diagonal matrix, and each element on the diagonal represents the thermal capacity of the corresponding node. Its value is equal to the product of the density, specific heat capacity, and volume of the node material. These parameters are derived from material handbooks and design drawings. The node thermal conduction matrix describes the heat conduction and heat convection capabilities between nodes. Its elements are calculated based on empirical formulas for material thermal conductivity, contact thermal resistance, and cooling wind speed. The initial values ​​are derived from design simulation. The ambient temperature vector comes from sensor data. The heat source distribution matrix is ​​responsible for mapping the volume loss density calculated by S2 to the corresponding nodes in the thermal network. By using numerical methods such as the backward Euler method to solve this set of differential equations, the transient three-dimensional temperature field inside the entire generator set can be obtained, and the highest temperature value, i.e., the hot spot temperature, can be identified. S4, the central processing module 4 combines the data from the insulation state sensing module 3 to estimate the dynamic state of insulation aging. Central processing module 4 calls the insulation aging kinetics model, whose instantaneous aging rate is determined by multiplying the thermal aging term and the electrical aging term; Input the hot spot temperature into the thermal aging term, which conforms to the Arrhenius equation form and characterizes the exponential effect of temperature on the chemical reaction rate. The total harmonic distortion rate of the output voltage is input into the electrical aging term, which is in the form of a power function and characterizes the accelerating effect of electrical stress on aging. The relative change in dielectric constant obtained from the insulation state sensing module 3 is used as a linear correction factor in the calculation of the aging rate. Integrating the aging rate over time yields the normalized cumulative aging degree of insulation, which varies continuously from zero to one, where one represents the end of life. More specifically, Based on both thermal and electrical stresses, the cumulative aging degree of insulating materials is assessed online. An insulation aging kinetic model is employed, in which the instantaneous aging rate is determined by multiplying the thermal aging term and the electrical aging term. The thermal aging term conforms to the Arrhenius equation, specifically calculated by dividing the activation energy (with a negative exponent) by the product of the ideal gas constant and the hot spot temperature, with the natural constant e as the base. The electrical aging term is calculated by dividing the total harmonic distortion rate of the output voltage by the reference harmonic distortion rate, and then raising the electrical aging stress exponent to the power of the exponent. The relative change in dielectric constant is derived from the measurement data of insulation state sensing module 3 and is used as a linear correction factor in the calculation. The normalized cumulative aging degree of insulation varies continuously from zero to one, where zero represents brand new and one represents the end of life. The end of life is usually defined as a percentage decrease in dielectric strength to its initial value. The pre-exponential factor is related to the microstructure of the material. The activation energy reflects the energy barrier required for the insulating material to undergo thermal aging. The ideal gas constant is a physical constant, and the hot spot temperature is derived from the calculation results of S3. The total harmonic distortion (THD) of the output voltage comes from the analysis in step two. The reference harmonic distortion is usually taken as the value under the rated pure sine wave condition. The electrical aging stress index characterizes the intensity of the influence of electric field or harmonics on aging. The dielectric constant feedback gain coefficient is an adjustment parameter. The initial values ​​of key parameters such as the exponential factor, activation energy, electrical aging stress index, reference harmonic distortion, and dielectric constant feedback gain coefficient cannot be directly measured. They need to be obtained by conducting accelerated aging tests on samples of the same type of insulation material, such as increasing the temperature or voltage, and then fitting the test data. During system operation, these parameters will be continuously calibrated by the subsequent S5. S5, the adaptive signal injection module 6 and the central processing module 4 work together to perform adaptive active detection and online calibration of model parameters; When the system evaluation uncertainty is high or the periodic calibration point is reached, the central processing module 4 instructs the adaptive signal injection module 6 to inject a set of test current signals with preset frequency and amplitude into the generator bus. The harmonic stress field reconstruction module 5 synchronously measures the voltage response of the generator set terminals and calculates the measured impedance frequency spectrum. Central processing module 4 inputs the current insulation aging degree and thermal network parameters into the theoretical model to generate the corresponding theoretical impedance frequency spectrum; With the goal of minimizing the difference between the measured spectrum and the theoretical spectrum, an optimization problem with regularization constraints is constructed. The key physical parameters in the insulation aging kinetic model and the thermal network digital model are solved in reverse and updated to reduce model uncertainty. More specifically, By actively injecting safe test signals into the generator set, the internal state of the system can be inferred, thereby calibrating potential errors in the model. First, a parameterized model response definition is established: when a set of test current signals with specific frequencies is injected, the voltage response at the generator set terminals can be measured, and the ratio of this voltage to the injected current is used to calculate the complex impedance, obtaining the measured impedance frequency spectrum. Then, based on the current thermal network parameters and material parameters affected by the normalized cumulative aging of the insulation, such as a potential decrease in thermal conductivity, the theoretical impedance frequency spectrum that the generator set should theoretically exhibit can be predicted through an electromagnetically and thermally coupled model. This prediction involves a set of key physical parameters to be calibrated, such as activation energy and electrical aging stress index. After considering factors such as the number of parameters and local thermal resistance, an optimization problem with regularization constraints is constructed: the goal is to find a set of parameter values ​​that minimizes the sum of squares of the differences between the theoretical impedance frequency spectrum and the measured impedance frequency spectrum at all injection frequency points. To avoid the model from over-fitting the current data and losing its generality, a regularization term is added to the optimization objective. This term is the square of the difference between the current parameter to be estimated and the prior parameter value multiplied by a regularization coefficient. By solving this optimization problem, a set of optimal parameters can be found, and these parameters can be used to update the insulation aging kinetic model in S4 and the thermal network digital model in S3, thereby reducing model uncertainty. Commonly used optimization algorithms include nonlinear least squares methods such as the Levenberg-Marquardt method. S6, Central Processing Module 4 performs comprehensive health assessment and dynamic prediction of remaining lifespan; Central processing module 4 subtracts the normalized cumulative aging degree from one to calculate the current health index; Central processing module 4 connects to or predicts the future load power and harmonic emission level curves of the generator set as stress input; The Monte Carlo simulation method is used to randomly generate a large number of sequences that conform to the future stress probability distribution, which drive the updated insulation aging dynamics model to iteratively calculate a large number of possible aging trajectories. The time taken for all aging trajectories to first reach the preset failure threshold is statistically analyzed to form the probability distribution of the remaining useful life, and its expected value and key quantile value are output as the prediction result. More specifically, The physical aging degree is transformed into an intuitive health indicator to predict the future service life of the generator set. The health index is obtained through a simple linear mapping, that is, subtracting the current normalized cumulative aging degree of the insulation from a number of one. When the health index is one, it represents a brand new state, and zero represents the end of the life. For the prediction of the remaining useful life, a conditional integral formula is used: the remaining life is equal to the reciprocal of the time required for each unit increment of aging degree in the process from the current aging degree integral to the aging degree limit value of one. The aging rate in the integral depends on the future stress conditions. The future stress profile includes information such as the predicted load power and harmonic emission level curves. This data can come from the load prediction system of the data center or the statistics of historical operating patterns. In the actual calculation, the Monte Carlo simulation method is used: according to the probability distribution of future stress, thousands of possible future stress sequences are randomly generated to drive the insulation aging dynamics model updated in step five to iteratively calculate forward, thereby obtaining thousands of possible aging trajectories. Once all these aging trajectories reach a preset failure threshold for the first time, i.e., the degree of aging equals the time elapsed, a probability distribution of the remaining useful life can be formed, and its expected value and key quantile values ​​can be output as the final prediction result. S7, the central processing module 4 and the closed-loop execution module 7 work together to generate and verify closed-loop risk mitigation decisions; Based on the predicted results of the health index and remaining useful life, the central processing module 4 compares them with preset thresholds and classifies the risk level into three levels: normal, warning, and critical. If the warning level is determined, the contribution weight of each harmonic to the temperature rise of the hot spot is analyzed, a targeted filtering instruction is generated, and sent to the active filter through the closed-loop execution module 7 to suppress high-contribution harmonics first. If the situation is determined to be critical, emergency start and shutdown maintenance commands are generated and sent to the automatic transfer switch and generator set controller through the closed-loop execution module 7. After executing the instruction, the system continuously monitors the rate of change of key parameters, calculates the strategy effectiveness index, and stores the complete "decision, response, and result" sequence in the case library for optimizing risk judgment rules and strategy generation logic. More specifically, Based on the risk assessment conclusions, the optimal maintenance or control strategy is generated and implemented to complete the entire management cycle. First, a harmonic contribution analysis is performed: the contribution weight of the h-th harmonic to the hotspot temperature rise is calculated, which is equal to the square of the harmonic current amplitude multiplied by the corresponding AC resistance, then divided by the total loss to obtain the proportion of the harmonic loss to the total loss. Finally, this is multiplied by the sensitivity coefficient of the hotspot temperature to the harmonic loss. The sensitivity coefficient of the hotspot temperature to a specific harmonic loss is obtained through disturbance analysis using the thermal network digital model in step three. Based on the health index and the predicted remaining useful life, the system executes preset decision rules, which are divided into multiple levels: when the health index is less than 0.3, or the probability of the remaining useful life being less than 24 hours exceeds 0.8, a critical state is determined. The system automatically starts the standby unit in parallel operation through the closed-loop execution module 7 and requests an emergency shutdown and maintenance of the main generator unit. When the health index is less than 0.6, or the expected value of the remaining useful life is less than 14... When the weather conditions are deemed to be in a warning state, the control action is to send a targeted filtering command to the active filter via the closed-loop execution module 7, causing it to adjust to targeted suppression mode. This prioritizes filtering out the first two to three harmonics that contribute the most to the temperature rise of the hotspot. The maintenance action is to automatically generate a work order, suggesting that during the next planned downtime, infrared checks and partial discharge verification be performed on the specific hotspot location located by S3. After executing the command, the system enters the learning and verification phase: after a preset period of time, it calculates the strategy effectiveness index, such as harmonic suppression efficiency, which is equal to the total harmonic distortion rate after the measure divided by the total harmonic distortion rate before the measure. At the same time, it calculates the temperature rise improvement rate, which is equal to the temperature rise before the measure minus the temperature rise after the measure, and then divided by the temperature rise before the measure. Finally, the complete decision response result sequence consisting of the health index of this event, the array of harmonic contribution weights, the adopted strategy, and the strategy effect is stored in the case library for future optimization of risk judgment rules and strategy generation logic.

[0038] To address the issues of winding overheating and insulation aging caused by harmonics, the remote monitoring system for diesel generator sets employs a dielectric constant sensor network 1 attached to the windings. This network monitors the dielectric constant of the insulating material—a physical quantity reflecting its molecular structure—and simultaneously monitors the composition of characteristic gases within the generator set. Furthermore, the harmonic stress field reconstruction module 5 utilizes high-precision synchronous sampling technology to continuously measure voltage and current waveforms. It then employs windowed interpolation and Fast Fourier Transform analysis to precisely decompose the amplitude and phase information of each specific harmonic, serving as the accurate starting point for all subsequent quantitative analyses. The adaptive signal injection module 6 injects fully controlled, specific-frequency test signals into the generator set when needed, enabling online active diagnostics.

[0039] To address the issue of uneven spatial distribution and inability to directly measure the additional heat generated by harmonics, a dynamic model is used to accurately convert the harmonic spectrum into a three-dimensional heat distribution map inside the generator set. Based on the precise harmonic data provided by the harmonic stress field reconstruction module 5, the central processing module 4, combined with the physical models of the generator set winding's three-dimensional structure and material thermal properties, calculates the additional loss density generated by each harmonic current at a specific location in the winding. Subsequently, this loss density is input as a heat source into the built-in three-dimensional thermal network digital model of the generator set. Taking cooling conditions into account, the transient three-dimensional temperature field inside the generator set is dynamically reconstructed by solving the thermal network differential equations, thereby accurately locating potential overheating areas and hotspot temperatures.

[0040] To address the challenge of quantifying and tracking the cumulative aging of insulating materials under thermal and electrical stress, an insulation aging kinetic model is established and updated online. The central processing module 4 inputs the calculated hotspot temperature, the measured total harmonic distortion rate of the output voltage, and the change in dielectric constant provided by the insulation state sensing module 3 into this mathematical model. The model describes thermal aging using the Arrhenius equation and electrical aging using a power function. Through integral calculations, it outputs a continuously varying normalized cumulative aging degree of the insulation from zero to one in real time, providing a clear quantification of the insulation's lifespan.

[0041] To address potential errors in the internal prediction model due to individual equipment variations and long-term aging, an adaptive active detection and online calibration mechanism is employed. When system evaluation uncertainty is high, the central processing module 4 instructs the adaptive signal injection module 6 to inject a test signal into the generator set, and the harmonic stress field reconstruction module 5 measures the response to obtain the measured impedance frequency spectrum. The system compares this measured spectrum with the theoretical impedance frequency spectrum predicted based on the current aging degree and thermal network parameters. By solving an optimization problem with regularization constraints, key physical parameters in the insulation aging kinetic model and the thermal network digital model are calibrated in reverse, ensuring that the model always closely approximates the actual physical state of the monitored generator set.

[0042] To proactively assess insulation condition and quantify failure risk, probabilistic lifetime prediction is performed based on an updated high-fidelity model. Central processing module 4 converts normalized cumulative aging into a health index. Combining future load and harmonic prediction curves of the data center, a Monte Carlo simulation method is used to randomly generate thousands of possible future stress sequences, driving iterative calculations of the aging model to obtain a large number of possible aging trajectories. The time it takes for these trajectories to reach the failure threshold is statistically analyzed to form a probability distribution of remaining useful lifetime, thereby outputting the expected lifetime and risk probability.

[0043] To automatically implement intervention measures based on predicted risks to break the vicious cycle, the central processing module 4, based on preset rules and a closed-loop decision-making and execution system, classifies risks into three levels: normal, warning, and critical, according to the health index and remaining lifespan probability distribution. If a warning is detected, the contribution weight of each harmonic to the hotspot temperature rise is analyzed, and a targeted filtering instruction is generated and sent to the active filter via the closed-loop execution module 7 to prioritize the suppression of high-contribution harmonics. If a critical level is detected, an instruction is generated to activate the standby unit in parallel and request an emergency shutdown. All execution effects are verified using a new round of sensing data, and the "decision, response, and result" sequence is stored in a case library for continuous optimization of the decision-making logic.

[0044] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A remote monitoring system for diesel generator sets, characterized in that, It includes a dielectric constant sensor network (1), a flexible capacitive sensor array (2), an insulation state sensing module (3), a harmonic stress field reconstruction module (5), an adaptive signal injection module (6), a central processing module (4), and a closed-loop execution module (7). The dielectric constant sensor network (1) and the flexible capacitor sensor array (2) are installed inside the diesel generator set. The dielectric constant sensor network (1) measures the capacitance change of the sensing unit embedded in the diesel generator set to reflect the molecular structure aging state of the insulating material. The flexible capacitor sensor array (2) monitors the physical state of the winding by sensing the capacitance change caused by the small deformation of the stator winding surface in the diesel generator set due to thermal stress or mechanical vibration. The insulation state sensing module (3) is connected to the dielectric constant sensor network (1) and the flexible capacitor sensor array (2). The insulation state sensing module (3) is connected to the central processing module (4). The voltage and current input channels of the harmonic stress field reconstruction module (5) are respectively connected to the secondary side of the voltage transformer and the current sensor in the diesel generator set, and its data output terminal is connected to the central processing module (4). Its synchronous clock input is connected to the system global clock source. The harmonic stress field reconstruction module (5) calculates the harmonic loss density distribution. The injection signal output terminal of the adaptive signal injection module (6) is used to connect to the output bus of the diesel generator set. Its response signal input terminal is coupled to the sensor signal link of the harmonic stress field reconstruction module (5). Its control and data interface is connected to the central processing module (4). The central processing module (4) performs dynamic reconstruction of the thermal stress field based on the harmonic loss density distribution, estimates the dynamic state of insulation aging by combining the data of the insulation state sensing module (3), and performs adaptive active detection and online calibration of model parameters in conjunction with the adaptive signal injection module (6). It also performs comprehensive health assessment and dynamic prediction of remaining life. The central processing module (4) is connected to the closed-loop execution module (7) to send control commands. The control signal output terminal of the closed-loop execution module (7) is connected to the active filter and control cabinet of the diesel generator set to receive and execute the commands from the central processing module (4).

2. The remote monitoring system for a diesel generator set according to claim 1, characterized in that, The dielectric constant sensor network (1) includes multiple flexible sensing units and a multiplexed data bus. Each flexible sensing unit includes a flexible substrate, interdigitated electrodes disposed on the flexible substrate, and an insulating protective layer covering the interdigitated electrodes. The leads of the two interdigitated electrodes of each flexible sensing unit are respectively connected to the signal line and the common reference ground line in the multiplexed data bus. The terminal of the multiplexed data bus is connected to the measurement input terminal of the capacitance digital converter in the insulation state sensing module (3).

3. The remote monitoring system for a diesel generator set according to claim 1, characterized in that, The flexible capacitive sensor array (2) includes multiple capacitive sensing units, a flexible flat cable and an integrated connector. Each capacitive sensing unit includes a flexible substrate, parallel plate electrodes disposed on the surface of the flexible substrate, and an elastomer coating layer encapsulating the parallel plate electrodes. The two parallel plate electrodes of each capacitive sensing unit are connected to the integrated connector through the flexible flat cable. The integrated connector is connected to the analog input channel of the capacitor-to-digital converter in the insulation state sensing module (3).

4. The remote monitoring system for a diesel generator set according to claim 1, characterized in that, The insulation state sensing module (3) includes a microcontroller, a capacitor-to-digital converter, a gas sensing signal conditioning circuit, a multi-channel analog-to-digital converter, and a first board-to-board connector. The digital interface of the capacitor-to-digital converter is connected to the microcontroller through a first digital isolator. The output of the gas sensing signal conditioning circuit is connected to the analog input channel of the multi-channel analog-to-digital converter. The digital output interface of the multi-channel analog-to-digital converter is connected to the serial peripheral interface of the microcontroller. The data input and data output of the first communication interface chip of the multi-channel analog-to-digital converter are connected to the transmit and receive ends of the microcontroller's universal asynchronous transceiver through a second digital isolator, respectively. The universal input / output group, the serial peripheral interface, and the differential output of the first communication interface chip of the microcontroller are all connected to the corresponding communication ends of the first board-to-board connector. The first board-to-board connector is connected to the central processing module (4).

5. The remote monitoring system for a diesel generator set according to claim 1, characterized in that, The harmonic stress field reconstruction module (5) includes a synchronous sampling analog-to-digital converter, a first signal conditioning circuit, a field-programmable gate array (FPGA), a phase-locked loop (PLL) circuit, an external signal connector, and a second board-to-board connector. The voltage input terminal and current input terminal of the external signal connector are respectively connected to the corresponding analog input channel of the synchronous sampling analog-to-digital converter through the first signal conditioning circuit. The reference clock input terminal of the PLL circuit is connected to the system global clock source, and its output clock terminal is connected to the external sampling clock input terminal of the synchronous sampling analog-to-digital converter and the global clock input terminal of the FPGA. The digital data output terminal of the synchronous sampling analog-to-digital converter is connected to the high-speed data input interface of the FPGA. The data bus interface, configuration interface, and general input / output terminal group of the FPGA are all connected to the second board-to-board connector. The second board-to-board connector is connected to the central processing module (4).

6. The remote monitoring system for a diesel generator set according to claim 1, characterized in that, The adaptive signal injection module (6) includes a direct digital frequency synthesizer, a digital-to-analog converter, a programmable gain amplifier, a power amplifier, an isolation transformer, a high-speed analog-to-digital converter, a second signal conditioning circuit, a directional coupler, a current limiting protection circuit, a transient voltage suppressor, and a third board-to-board connector; the parallel data output port of the direct digital frequency synthesizer is connected to the parallel data input port of the digital-to-analog converter; the analog voltage output port of the digital-to-analog converter is connected in series with the input terminal of the programmable gain amplifier, the input terminal of the power amplifier, and the primary winding of the isolation transformer; the secondary winding of the isolation transformer is connected to... After the current limiting protection circuit and the transient voltage suppressor, the module is connected to the generator output bus through a dedicated interface on the front panel for injecting test signals; the coupling end of the directional coupler is connected in series in the secondary signal link of the generator current transformer, and its output end is connected in sequence to the input end of the second signal conditioning circuit and the analog input channel of the high-speed analog-to-digital converter for acquiring the injected response signal; the reference clock input end of the direct digital frequency synthesizer, the configuration bus interface, and the data output interface of the high-speed analog-to-digital converter are all connected to the third board-to-board connector, which is connected to the central processing module (4).

7. The remote monitoring system for a diesel generator set according to claim 1, characterized in that, The central processing module (4) includes a multi-core processor, a field-programmable gate array (FPGA), a time-sensitive network (TSN) Ethernet switch chip, a non-volatile memory, a watchdog circuit, and a fourth board-to-board connector. The multi-core processor is connected to the FPGA and the TSN Ethernet switch chip via a high-speed serial computer expansion bus standard channel. The multiple general-purpose input / output (GPIO) terminals and the high-speed serial interface of the FPGA are all connected to the fourth board-to-board connector. The non-volatile memory is connected to the multi-core processor via a serial peripheral interface or a parallel bus. The reset signal output terminal of the watchdog circuit is connected to the system reset terminal of the multi-core processor. The fourth board-to-board connector is connected to the TSN Ethernet switch chip and the FPGA's GPIO terminals.

8. The remote monitoring system for a diesel generator set according to claim 1, characterized in that, The closed-loop execution module (7) includes an isolated digital output chip, an isolated analog output chip, an optocoupler isolator, a power driver, a digital signal isolator, a communication protocol conversion chip, and a fifth board-to-board connector. The data input port of the isolated digital output chip is connected to the data bus terminal of the fifth board-to-board connector through the digital signal isolator. The digital control interface of the isolated analog output chip is connected to another output channel of the digital signal isolator. Each output channel of the isolated digital output chip is connected to the first set of terminals of the terminal block through the optocoupler isolator and the power driver in sequence. The current output port of the isolated analog output chip is connected to the second set of terminals of the terminal block. The terminal block is used to connect the control input terminal of the active filter and the remote control terminal of the control cabinet in the generator set. One end of the communication protocol conversion chip is connected to the auxiliary communication terminal of the fifth board-to-board connector through a serial bus. The fifth board-to-board connector is connected to the central processing module (4).

Images