320 cylinder diameter dual-fuel engine electric control system and method based on multi-mode cooperative control

Abstract

According to the technical scheme, the 320-cylinder-diameter dual-fuel engine electric control system is characterized by comprising a main MCU control module and an FPGA real-time monitoring module which are connected with each other, the FPGA real-time monitoring module is connected with a security module, the security module is connected with a fuel gas leakage monitoring module, and meanwhile, the FPGA real-time monitoring module is connected with the fuel gas leakage monitoring module. A data receiving end of the security module is connected with the sensor, the ESD button and the oil mist detector, and a control signal output end of the security module is connected with the GVU pressure regulating valve, the exhaust valve, the alarm device and the fuel oil stop valve; the FPGA real-time monitoring module is connected into the Ethernet and the hard wiring harness, the pressure sensor, the temperature sensor, the liquid level switch, the on-site remote change-over switch, the combustion analyzer, the waste gas bypass valve, the gas valve, the main lubricating oil pump, the exhaust ventilation valve, the exhaust valve, other auxiliary equipment, the cartridge valve, the gas adjusting valve and buttons of the LOP module are connected into the hard wiring harness, and a display screen of the LOP module is connected into the Ethernet. According to the other technical scheme, the invention provides an electronic control method for the 320 cylinder diameter dual-fuel engine based on multi-mode cooperative control.

Description

Technical Field

[0001] The present invention relates to an electronic control system and method for a 320 cylinder diameter dual-fuel engine based on multi-modal coordinated control. Background Art

[0002] Currently, large-bore marine engines operate under complex conditions, making it difficult for traditional electronic control systems to achieve high-precision coordinated dual-fuel control. This leads to problems such as unstable combustion, low thermal efficiency, and excessive emissions. Furthermore, large-bore marine engines exhibit poor dynamic response during fuel switching, which can easily lead to power interruptions or detonation. Existing control systems lack adaptability to multiple environmental parameters (air pressure, humidity, and fuel composition). Summary of the Invention

[0003] The purpose of the present invention is that the operating conditions of large-cylinder marine engines are complex, and traditional electronic control systems are difficult to achieve high-precision coordinated control of dual fuels.

[0004] In order to achieve the above-mentioned purpose, a technical solution of the present invention is to disclose an electronic control system for a 320-cylinder dual-fuel engine based on multi-modal collaborative control, which is characterized in that it includes a main MCU control module and an FPGA real-time monitoring module connected to each other, the FPGA real-time monitoring module is connected to the security module, and the security module is connected to the gas leakage monitoring module. At the same time, the data receiving end of the security module is connected to the sensor, ESD button and oil mist detector, and the control signal output end of the security module is connected to the GVU pressure regulating valve, exhaust valve, alarm device and fuel shut-off valve; the FPGA real-time monitoring module is connected to Ethernet and a hard-wired wiring harness, and the pressure sensor, temperature sensor, liquid level switch, local remote switching switch, combustion analyzer, exhaust bypass valve, gas valve, main lubricating oil pump, exhaust ventilation valve, exhaust valve, other auxiliary equipment, cartridge valve, gas regulating valve and buttons of the local operation cabinet are connected to the hard-wired wiring harness, and the display screen of the local operation cabinet is connected to Ethernet.

[0005] Preferably, the main MCU control module and the FPGA real-time monitoring module are connected via a CAN bus.

[0006] Preferably, the FPGA real-time monitoring module is connected to the security module via an RS485 bus.

[0007] Preferably, the security module is connected to the gas leakage monitoring module via an RS485 bus.

[0008] Another technical solution of the present invention is to provide an electronic control method for a 320 cylinder diameter dual-fuel engine based on multi-modal coordinated control, characterized in that it includes the following steps:

[0009] Step 1: Input real-time data including cylinder pressure waveform, exhaust temperature, NOx concentration, and fuel injection amount into the data acquisition module through sensors, and perform data preprocessing through filtering and normalization;

[0010] Step 2: The feature extraction module performs the following processing:

[0011] Signal processing: FFT analysis of cylinder pressure waveform to extract combustion oscillation frequency characteristics;

[0012] Statistical features: Sliding window calculation of mean / variance;

[0013] Temporal correlation: using LSTM networks to capture the temporal dependencies of multiple parameters;

[0014] The feature extraction module splices the extracted signals into a multi-dimensional feature vector;

[0015] Step 3: Update the model, specifically:

[0016] The model inputs are the feature vector and the target variable. Online training is performed using an incremental learning method to update the neural network weights. Finally, the predicted control parameters are compared with the actual combustion results to verify the model.

[0017] Step 4: Use LSTM neural network and random forest regression to complete the input of the model, conduct verification on the model, and evaluate the combustion stability indicators.

[0018] Compared with the existing technical solutions, the present invention has the following beneficial effects:

[0019] 1. Dual-fuel coordinated control strategy and dynamic fuel ratio algorithm: Based on engine speed, load, emission data and fuel characteristics (such as methane number and methanol concentration), the optimal fuel mixture ratio is calculated in real time. Multi-mode switching logic: The three-level mode of "diesel-dominant, dual-fuel, and gas-dominant" is preset, and smooth transition is achieved through fuzzy PID control;

[0020] 2. Multi-parameter closed-loop control system, sensor fusion: Integrate cylinder pressure sensor, knock sensor, NOx / O2 sensor to build a real-time feedback network for combustion status;

[0021] 3. Injection timing optimization: Feedforward-feedback composite control is used to dynamically adjust the phase difference between the diesel pilot injection and the gas main injection to suppress knock;

[0022] 4. Adaptive learning module, online parameter calibration: Through the neural network model (LSTM) to learn historical operation data, dynamically optimize control parameters (EGR rate, ignition advance angle);

[0023] 5. Fuel quality self-identification: Based on combustion pressure waveform analysis, it automatically identifies fuel calorific value deviation and compensates for it;

[0024] 6. Redundant fault-tolerant mechanism, dual CAN bus architecture: the main control unit (MCU) and the backup unit (FPGA) run in parallel, seamless switching in case of failure;

[0025] 7. Fault diagnosis tree: A built-in rule-based fault diagnosis system supports rapid response to scenarios such as fuel leakage and sensor failure. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 This is the overall system architecture block diagram. In the figure, LOP stands for local operation cabinet, including local operation buttons and display screen;

[0027] Figure 2 This is the flow chart of the adaptive learning module. DETAILED DESCRIPTION

[0028] Below in conjunction with specific embodiment, further set forth the present invention.Should be understood that these embodiments are only used to illustrate the present invention and are not used in limiting the scope of the present invention.In addition, should be understood that after reading the content taught by the present invention, those skilled in the art can make various changes or modifications to the present invention, and these equivalent forms fall equally within the scope limited by the appended claims of the application.

[0029] like Figure 1As shown, the present invention provides an electronic control system for a 320-cylinder dual-fuel engine, comprising a main MCU control module and a FPGA real-time monitoring module. The main MCU control module and the FPGA real-time monitoring module are connected via a CAN bus, and the FPGA real-time monitoring module is connected to a security module via an RS485 bus. The security module is connected to a gas leak monitoring module via the RS485 bus. The security module also receives data from sensors, ESD buttons, and oil mist detectors. It also sends control signals to the GVU pressure regulating valve, exhaust valve, alarm device, and fuel shut-off valve. The FPGA real-time monitoring module is connected to Ethernet and a hard-wiring harness. The pressure sensor, temperature sensor, liquid level switch, local remote selector switch, combustion analyzer, wastegate valve, gas valve, main oil pump, exhaust ventilation valve, exhaust valve, other auxiliary equipment, cartridge valves, gas regulating valve, and buttons of the local operation panel (LOP) are connected to the hard-wiring harness. The display screen of the local operation panel (LOP) is connected to Ethernet. The local operation cabinet has physical display lights: power indicator, local display, operation display, parking display, emergency operation, vehicle preparation completion, gas mode, fuel mode and sound and light alarm devices, etc.; there are operation buttons: emergency stop, local / remote control, start / stop, mute, reset, vehicle preparation, speed increase / deceleration, start / blow down, emergency start, oil and gas switch, GVU start, etc.; the screen on the LOP cabinet displays the speed, load and temperature and pressure signals on the engine

[0030] System redundant controller work switching process:

[0031] Normal operation (i.e. master control mode): controlled by the main MCU control module and monitored in real time by the FPGA real-time monitoring module;

[0032] During fault detection: When a sensor fails or a communication anomaly is detected, the FPGA real-time monitoring module takes over control;

[0033] System recovery: After the main MCU control module restarts, data is synchronized;

[0034] Shutdown protection state: safe shutdown under serious fault.

[0035] In this embodiment, the redundant controller hardware architecture is as follows:

[0036] The main MCU control module adopts a multi-core MCU (model: Infineon Aurix TC397), and the FPGA real-time monitoring module adopts FPGA (model: Xilinx Zynq UltraScale+).

[0037] In this embodiment, the software process of the above system includes the following steps:

[0038] 1) Initialization phase:

[0039] Load the fuel characteristic parameter library (such as natural gas composition table).

[0040] 2) Operation phase:

[0041] Cylinder pressure and exhaust data are collected in each cycle to calculate IMEP (Indicated Mean Effective Pressure) and combustion heat release rate.

[0042] The sensor noise is eliminated through the Kalman filter and input to the control algorithm.

[0043] Output injection pulse width and boost pressure setting value to the actuator.

[0044] 3) Troubleshooting:

[0045] Trigger graded alarm (early warning - power reduction - shutdown).

[0046] In the embodiment of the present invention, the flowchart of the system processing the adaptive learning module is as follows: Figure 2 As shown;

[0047] 1. Input real-time data such as cylinder pressure waveform, exhaust temperature, NOx concentration, and fuel injection amount into the data acquisition module through sensors; perform data preprocessing through filtering and normalization;

[0048] 2. Feature extraction module:

[0049] Signal processing: FFT analysis of cylinder pressure waveform is used to extract combustion oscillation frequency characteristics.

[0050] Statistical features: Sliding window calculation of mean / variance (such as the fluctuation characteristics of NOx concentration).

[0051] Temporal Correlation: Using LSTM networks to capture the temporal dependencies of multiple parameters.

[0052] The extracted signals are spliced into multi-dimensional feature vectors.

[0053] 3. Model update module:

[0054] Model input: feature vector + target variable (such as optimal EGR rate, ignition advance angle).

[0055] Online training: Incremental learning updates the neural network weights.

[0056] Verification mechanism: Comparison of predicted control parameters with actual combustion results (IMEP deviation).

[0057] 4. Use LSTM neural network and random forest regression to complete the model input, conduct verification on the model, and evaluate combustion stability indicators (such as speed fluctuation rate).

Claims

1. An electronic control system for a 320 cylinder diameter dual-fuel engine based on multi-modal coordinated control, characterized in that: It includes an interconnected main MCU control module and an FPGA real-time monitoring module. The FPGA real-time monitoring module is connected to the security module, and the security module is connected to the gas leakage monitoring module. At the same time, the data receiving end of the security module is connected to the sensor, ESD button and oil mist detector, and the control signal output end of the security module is connected to the GVU pressure regulating valve, exhaust valve, alarm device and fuel shut-off valve; the FPGA real-time monitoring module is connected to Ethernet and a hard-wired wiring harness, and the pressure sensor, temperature sensor, liquid level switch, local remote switching switch, combustion analyzer, exhaust gas bypass valve, gas valve, main lubricating oil pump, exhaust ventilation valve, exhaust valve, other auxiliary equipment, plug-in valve, gas regulating valve and buttons of the local operation cabinet are connected to the hard-wired wiring harness, and the display screen of the local operation cabinet is connected to Ethernet.

2. The electronic control system for a 320 cylinder diameter dual-fuel engine based on multi-modal coordinated control according to claim 1, characterized in that: The main MCU control module is connected to the FPGA real-time monitoring module via a CAN bus.

3. The electronic control system for a 320 cylinder diameter dual-fuel engine based on multi-modal coordinated control according to claim 1, characterized in that: The FPGA real-time monitoring module is connected to the security module via an RS485 bus.

4. The electronic control system for a 320 cylinder diameter dual-fuel engine based on multi-modal coordinated control according to claim 1, characterized in that: The security module is connected to the gas leakage monitoring module via an RS485 bus.

5. An electronic control method for a 320 cylinder diameter dual-fuel engine based on multi-modal coordinated control, characterized in that: The following steps are involved: Step 1: Input real-time data including cylinder pressure waveform, exhaust temperature, NOx concentration, and fuel injection amount into the data acquisition module through sensors, and perform data preprocessing through filtering and normalization; Step 2: The feature extraction module performs the following processing: Signal processing: FFT analysis of cylinder pressure waveform to extract combustion oscillation frequency characteristics; Statistical features: Sliding window calculation of mean / variance; Temporal correlation: using LSTM networks to capture the temporal dependencies of multiple parameters; The feature extraction module splices the extracted signals into a multi-dimensional feature vector; Step 3: Update the model, specifically: The model inputs are the feature vector and the target variable. Online training is performed using an incremental learning method to update the neural network weights. Finally, the predicted control parameters are compared with the actual combustion results to verify the model. Step 4: Use LSTM neural network and random forest regression to complete the input of the model, conduct verification on the model, and evaluate the combustion stability indicators.

Images