Intelligent cooperative emission control system applied to heavy vehicle type

Abstract

The invention provides an intelligent collaborative emission control system applied to heavy vehicles, and belongs to the technical field of heavy vehicle emission control, and the system comprises a data sensing unit, a data interaction unit, a cloud intelligent decision module, an instruction execution unit and an online diagnosis and fault tolerance unit. Operation and emission data are collected through sensors arranged on an engine and a post-processing system, and are uploaded to a cloud platform through vehicle-mounted communication; the cloud end predicts the emission trend based on the LSTM neural network, generates a dynamic control strategy in combination with a multi-objective optimization algorithm, and automatically matches local emission regulations according to the geographic position of the vehicle; the instruction execution unit is used for cooperatively regulating and controlling an engine, a crankcase ventilation valve and a fuel evaporation system; and the online diagnosis unit monitors the state of the system in real time to realize fault diagnosis and fault-tolerant control. According to the invention, intelligent, self-adaptive and compliant control of emission under all working conditions can be realized.

Description

Technical Field

[0001] This invention relates to the field of heavy-duty vehicle emission control technology, and in particular to an intelligent collaborative emission control system for heavy-duty vehicles. Background Technology

[0002] Heavy-duty vehicles are a major source of emissions from road traffic, making their emission control a key focus in environmental protection. Current emission control systems for heavy-duty vehicles are mostly based on fixed strategies using local ECUs, lacking the ability to coordinate and optimize for multiple pollutants and failing to adapt to complex and changing road conditions and regulatory requirements in different regions. Traditional systems often focus on controlling a single pollutant, making it difficult to achieve comprehensive emission reductions of nitrogen oxides, particulate matter, and carbon dioxide while ensuring power output. Furthermore, existing systems lack predictive and forward-looking control capabilities based on real-time data, and their fault diagnosis and tolerance mechanisms are inadequate, resulting in unstable emission control performance, especially under low-load, high-emission-risk conditions. Therefore, there is an urgent need for an emission control system that can intelligently coordinate multiple emission control devices, support multi-objective optimization, and possess online diagnostics and regulatory adaptive capabilities. Summary of the Invention

[0003] Therefore, the purpose of this invention is to provide an intelligent collaborative emission control system for heavy-duty vehicles, so as to at least solve the above problems.

[0004] The technical solution adopted in this invention is as follows: An intelligent collaborative emission control system for heavy-duty vehicles, the system comprising: The data sensing unit, which is deployed in the vehicle engine and after-treatment system, is used to collect operating status signals including engine speed, load, intake air temperature, exhaust air temperature, crankcase internal pressure, fuel vapor canister pressure and saturation, as well as emission signals such as nitrogen oxide concentration, particulate matter number and volatile organic compound concentration through emission sensors. The data interaction unit is communicatively connected to the data sensing unit and is used to package and upload the encapsulated sensing data to the cloud service platform through the vehicle wireless communication module, and to receive downlink control commands from the cloud service platform. The cloud-based intelligent decision-making module, deployed on a cloud service platform, includes: The emission prediction unit uses a prediction model based on a long short-term memory neural network to perform time-series analysis on the received real-time sensor data in order to predict the concentration change trend of each emission within a specific future time window. The collaborative optimization unit is used to optimize and solve the dynamic control strategy based on the output of the emission prediction unit, combined with the current engine operating point and the working status of the aftertreatment device, with the goal of simultaneously minimizing nitrogen oxide emissions, particulate matter emissions and carbon dioxide emissions. The regulatory library management unit stores limit charts and test procedures for multiple sets of emission regulations, and can automatically select the applicable regulatory version based on vehicle location information or user input, providing boundary constraints for the collaborative optimization unit. The command execution unit is connected to the engine electronic control unit, the selective catalytic reduction system urea injection controller, the diesel particulate filter regeneration controller, the electronic crankcase ventilation valve, and the fuel evaporation solenoid valve. It is used to receive and parse the control commands issued by the cloud-based intelligent decision module and drive the corresponding actuators to make adjustments. The online diagnostic and fault-tolerant unit interacts with the data sensing unit and the instruction execution unit to monitor the working status of the crankcase ventilation system and the fuel evaporation system in real time, perform fault diagnosis based on the comparison of sensor readings with preset thresholds, and activate a predefined adaptive compensation strategy when an anomaly is detected.

[0005] Furthermore, the data sensing unit includes a high-precision nitrogen oxide sensor, a particulate matter number counter, a photoionization volatile organic compound sensor, a piezoresistive crankcase pressure sensor, and a fuel vaporization carbon canister with integrated pressure and temperature sensing functions.

[0006] Furthermore, the input parameter set of the emission prediction unit includes at least engine speed, engine output torque, engine intake manifold temperature, upstream exhaust temperature of the selective catalytic reduction system, and cumulative urea injection quantity; the prediction model is jointly trained by bench tests and actual road emission test data, and can predict whether emissions are at risk of exceeding standards several seconds in advance.

[0007] Furthermore, the collaborative optimization unit uses the Pareto front optimization method to solve the multi-objective optimization problem and has a built-in dynamic weight allocator. This weight allocator can automatically adjust the weight coefficients of different emission optimization objectives according to the current operating conditions of the vehicle. The operating conditions include at least urban congestion conditions, suburban cruising conditions, and high-speed acceleration conditions.

[0008] Furthermore, the control actions generated by the command execution unit include adjusting the engine ignition angle and injection parameters, correcting the urea injection MAP, triggering the active regeneration of the diesel particulate filter, adjusting the duty cycle signal of the electronic crankcase ventilation valve to change the opening degree, and controlling the opening timing and duration of the fuel evaporation solenoid valve.

[0009] Furthermore, in the online diagnostics and fault-tolerant unit, When the internal pressure of the crankcase is continuously higher than the first pressure threshold, it is determined that the crankcase ventilation valve is blocked and a first alarm signal is generated. When the pressure in the fuel vapor carbon canister is continuously lower than the second pressure threshold, it is determined that there is a leak in the evaporation system and a second alarm signal is generated. When the calculated carbon canister saturation value exceeds the capacity threshold, a request is sent to the instruction execution unit to start the carbon canister desorption and regeneration process.

[0010] Furthermore, it also includes an engine bench calibration database, which stores fuel consumption and emission characteristic maps covering the entire operating condition universal characteristic range obtained through engine bench tests; the collaborative optimization unit calls the data in this database as a basic reference when generating dynamic control strategies, and executes a collaborative strategy to delay or reduce urea injection to suppress ammonia escape under low engine load conditions.

[0011] Furthermore, the regulatory database management unit is used to receive geographic location information sent from the vehicle-mounted global positioning system receiver or the vehicle-to-everything (V2X) communication module, and automatically match and load regulatory limit files that meet the current emission requirements of the region based on the location information, thereby realizing automatic switching of the applicable regulatory areas.

[0012] Furthermore, when the online diagnostic and fault-tolerant unit generates the first alarm signal or the second alarm signal, it sends a degraded operation command to the command execution unit. The command execution unit then controls the engine electronic control unit to enter the torque limiting mode and stores the corresponding fault code in the on-board diagnostic system.

[0013] Compared with the prior art, the beneficial effects of the present invention are: This invention achieves coordinated control of nitrogen oxides, particulate matter, and carbon dioxide emissions from heavy-duty vehicles through cloud-based intelligent decision-making and multi-objective optimization algorithms, significantly improving overall emission performance. It employs neural networks for emission trend prediction, providing forward-looking control capabilities to mitigate the risk of exceeding emission standards. A regulatory database management unit supports automatic switching between regulations from multiple regions, enhancing the system's applicability and compliance. Integrated online diagnostics and fault-tolerance mechanisms enable real-time monitoring of crankcase ventilation and fuel evaporation system status, timely fault detection, and the implementation of degraded operation strategies to ensure system safety and controllable emissions. Furthermore, by combining an engine bench calibration database, it optimizes urea injection strategies under low-load conditions, effectively suppressing ammonia escape and achieving coordinated emission reduction of multiple pollutants. Attached Figure Description

[0014] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only preferred embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0015] Figure 1This is a schematic diagram of the overall structure of an intelligent collaborative emission control system for heavy-duty vehicles proposed in an embodiment of the present invention. Detailed Implementation

[0016] The principles and features of the present invention are described below with reference to the accompanying drawings. The listed embodiments are only used to explain the present invention and are not intended to limit the scope of the present invention.

[0017] Reference Figure 1 This invention provides an intelligent collaborative emission control system for heavy-duty vehicles, the system comprising: The data sensing unit, which is deployed in the vehicle engine and after-treatment system, is used to collect operating status signals including engine speed, load, intake air temperature, exhaust air temperature, crankcase internal pressure, fuel vapor canister pressure and saturation, as well as emission signals such as nitrogen oxide concentration, particulate matter number and volatile organic compound concentration through emission sensors. The data interaction unit is communicatively connected to the data sensing unit and is used to package and upload the encapsulated sensing data to the cloud service platform through the vehicle wireless communication module, and to receive downlink control commands from the cloud service platform. The cloud-based intelligent decision-making module, deployed on a cloud service platform, includes: The emission prediction unit uses a prediction model based on a long short-term memory neural network to perform time-series analysis on the received real-time sensor data in order to predict the concentration change trend of each emission within a specific future time window. The collaborative optimization unit is used to optimize and solve the dynamic control strategy based on the output of the emission prediction unit, combined with the current engine operating point and the working status of the aftertreatment device, with the goal of simultaneously minimizing nitrogen oxide emissions, particulate matter emissions and carbon dioxide emissions. The regulatory library management unit stores limit charts and test procedures for multiple sets of emission regulations, and can automatically select the applicable regulatory version based on vehicle location information or user input, providing boundary constraints for the collaborative optimization unit. The command execution unit is connected to the engine electronic control unit, the selective catalytic reduction system urea injection controller, the diesel particulate filter regeneration controller, the electronic crankcase ventilation valve, and the fuel evaporation solenoid valve. It is used to receive and parse the control commands issued by the cloud-based intelligent decision module and drive the corresponding actuators to make adjustments. The online diagnostic and fault-tolerant unit interacts with the data sensing unit and the instruction execution unit to monitor the working status of the crankcase ventilation system and the fuel evaporation system in real time, perform fault diagnosis based on the comparison of sensor readings with preset thresholds, and activate a predefined adaptive compensation strategy when an anomaly is detected.

[0018] For example, the data sensing unit is deployed on the vehicle's engine and aftertreatment system, responsible for collecting two types of signals: operating status signals and emission signals. Operating status signals include engine speed, load, intake air temperature, exhaust air temperature, crankcase internal pressure, and the pressure and saturation of the fuel vapor canister. These parameters reflect the real-time operating status of the engine and auxiliary systems. Emission signals can be directly measured using dedicated emission sensors to measure the concentration of nitrogen oxides, particulate matter, and volatile organic compounds in the exhaust gas. The data interaction unit can receive sensor data from the data sensing unit, encapsulate and package it, and upload the data to a cloud service platform via an onboard wireless communication module. It can also receive control commands from the cloud service platform and transmit them to local actuators. The cloud-based intelligent decision-making module is deployed on a remote cloud service platform and comprises three sub-units: an emission prediction unit, a collaborative optimization unit, and a regulatory library management unit. The emission prediction unit uses a prediction model based on a long short-term memory neural network to perform time-series analysis on uploaded real-time sensor data, predicting the changing trends of various emission concentrations over a specific future period. The collaborative optimization unit, based on the prediction results of the emission prediction unit, comprehensively considers the engine's current operating point and the operating status of the aftertreatment device. Its goal is to simultaneously minimize nitrogen oxide emissions, particulate matter emissions, and carbon dioxide emissions by solving a multi-objective optimization problem, generating a dynamic control strategy. The regulatory library management unit stores various emission regulation limits and test procedures. Based on real-time vehicle location information or user input, the regulatory library management unit can automatically select and apply the currently applicable regulatory version, providing the collaborative optimization unit with the necessary regulatory boundaries and constraints. The command execution unit is located locally in the vehicle, and the regulatory library management unit is directly connected to multiple key controllers and control valves, including the engine electronic control unit, the urea injection controller of the selective catalytic reduction system, the regeneration controller of the diesel particulate filter, the electronic crankcase ventilation valve, and the fuel evaporation solenoid valve. Its function is to receive and parse control commands from the cloud-based intelligent decision-making module, and then drive the corresponding actuators to make adjustments.

[0019] The online diagnostic and fault-tolerant unit interacts with the data sensing unit and the command execution unit to monitor the working status of the crankcase ventilation system and the fuel evaporation system in real time. By comparing the sensor readings with preset safety thresholds, it can perform fault diagnosis. Once an abnormality is detected, it will activate a pre-set adaptive compensation strategy to maintain the basic functions of the system or enter a safe operating mode.

[0020] This system utilizes cloud computing power for AI-based emission prediction and multi-objective collaborative optimization, and closely integrates regulatory requirements with real-time fault diagnosis, thereby achieving intelligent, adaptive, and compliant emission control for heavy-duty vehicles under all operating conditions.

[0021] The data sensing unit includes a high-precision nitrogen oxide sensor, a particulate matter number counter, a photoionization volatile organic compound sensor, a piezoresistive crankcase pressure sensor, and a fuel vapor canister with integrated pressure and temperature sensing functions.

[0022] For example, the data sensing unit includes five types of sensors: a high-precision nitrogen oxide sensor, which directly measures the concentration of nitrogen oxides in engine exhaust, providing core data for emission control and assessment; a particulate matter number counter, which accurately counts the number of suspended particulate matter in exhaust gas, used to monitor particulate matter emission levels; a photoionization volatile organic compound sensor, which uses high-energy photons to ionize gas molecules, thereby detecting and measuring the concentration of volatile organic compounds in exhaust gas; a piezoresistive crankcase pressure sensor, which operates based on the piezoresistive effect, and can sensitively monitor changes in internal crankcase pressure, used to diagnose the status of the ventilation system; and a fuel vapor canister with integrated pressure and temperature sensing functions, which can simultaneously monitor pressure changes and temperature conditions inside the canister and calculate the saturation of fuel vapor.

[0023] The input parameter set of the emission prediction unit includes at least engine speed, engine output torque, engine intake manifold temperature, upstream exhaust temperature of the selective catalytic reduction system, and cumulative urea injection quantity; the prediction model is jointly trained by bench tests and actual road emission test data, and can predict whether emissions are at risk of exceeding standards several seconds in advance.

[0024] For example, the input parameter set of the emission prediction unit should at least include engine speed, engine output torque, engine intake manifold temperature, exhaust temperature upstream of the selective catalytic reduction system, and cumulative urea injection quantity. These parameters together constitute the basic data set for the prediction model's analysis and calculation. Engine speed and output torque: These two parameters define the engine's current operating point. Under different operating points, the engine's combustion efficiency, combustion temperature, and pollutant formation mechanisms (such as the close relationship between nitrogen oxide formation and temperature) show significant differences. Engine intake manifold temperature: Intake temperature affects intake density and charging efficiency, which in turn indirectly affects the air-fuel ratio and combustion temperature of the combustion process, and is an important precursor signal for emission prediction. Exhaust temperature upstream of the selective catalytic reduction system: This is a key limiting factor for the efficiency of the aftertreatment system. The selective catalytic reduction system needs to operate efficiently within a specific temperature window; if the temperature is too low, the catalytic reaction efficiency is low, and if the temperature is too high, it may lead to catalyst sintering failure. This temperature is the core input for predicting nitrogen oxide conversion efficiency. Cumulative urea injection volume: This parameter reflects the historical total amount of reducing agent (urea aqueous solution) that has been added to reduce nitrogen oxides. It helps the model estimate the storage level of ammonia in the selective catalytic reduction system and is a key state variable for predicting the future risk of "ammonia leakage" (i.e. ammonia escape) and the sustainability of nitrogen oxide conversion capacity.

[0025] The model is jointly trained using bench test data and real-world emissions test data. Bench test data, obtained in a controlled laboratory environment, provides high-precision, complete emissions data covering the full operating range of the engine, serving as the basic framework and initial accuracy of the model. Real-world emissions test data supplements the complex factors of the real world that bench data cannot cover, such as varying driving behaviors, actual ambient temperature and altitude, and different road gradients. Combining these two methods for joint training significantly improves the model's generalization ability and real-world predictive accuracy, preventing it from becoming a mere "theoretical model" applicable only to laboratory environments.

[0026] "Several seconds in advance" indicates that this is a forward-looking predictive control rather than a reactive response. This provides a valuable decision-making and execution time window for the downstream "cooperative optimization unit" to take optimization measures (such as adjusting fuel injection, urea injection, etc.), enabling the system to prevent problems before they occur.

[0027] "Exceeding the standard risk" means that the prediction results will be compared with the limits provided by the "regulatory database management unit" in real time. The system focuses not only on the absolute value of emissions, but also on their compliance margin relative to the regulatory limits, thereby achieving proactive compliance management.

[0028] The collaborative optimization unit uses the Pareto front optimization method to solve multi-objective optimization problems and has a built-in dynamic weight allocator. This weight allocator can automatically adjust the weight coefficients of different emission optimization objectives according to the current operating conditions of the vehicle. The operating conditions include at least urban congestion conditions, suburban cruising conditions, and high-speed acceleration conditions.

[0029] For example, the Pareto front optimization method does not seek a "unique optimal" solution, but rather strives to create an "optimal trade-off map." Each point on this map represents a possible control strategy, such as strategy A (extremely low NOx, but slightly high PM), strategy B (balanced PM and NOx, lowest CO2), and strategy C (extremely low CO2, but relatively high NOx and PM). All these points share a common characteristic: without worsening at least one other objective, no objective can be further improved. The boundary formed by these points is called the "Pareto front." After generating the Pareto front, the dynamic weight allocator needs to select the most suitable final strategy from this set of "non-dominated solutions" for the current situation. The system continuously analyzes data streams from the data sensing unit, such as vehicle speed, acceleration, throttle opening, and GPS location, to accurately determine the current driving condition of the vehicle. For each preset typical condition, the system has a set of predefined optimization priority rules. When identified as urban congestion: Frequent vehicle starts and stops lead to incomplete combustion. The primary concern here is the generation of large amounts of nitrogen oxides and particulate matter, which are directly harmful to human health. Therefore, the weighting allocator automatically and significantly increases the weighting coefficients of NOx and PM optimization targets. The instruction system prioritizes control strategies that maximally suppress the emissions of these two pollutants, even if this slightly increases carbon dioxide emissions (fuel consumption). When identified as suburban cruising: The engine runs smoothly within its high-efficiency range. The primary concern here is reducing operating costs and mitigating the greenhouse effect. Therefore, the weighting allocator reduces the weights of NOx and PM while significantly increasing the weight of CO2 optimization targets. The instruction system selects the strategy with the best fuel economy to achieve energy saving and emission reduction during long-distance driving. When identified as high-speed acceleration: The driver has a clear power demand. At this point, the weight allocator will adjust the weights appropriately while ensuring that emissions do not exceed the limits, so as to ensure that power response is not excessively sacrificed and to achieve a balance between emission control and driving performance. Finally, this weight combination, which is dynamically calculated based on real-time operating conditions, is used to select the most appropriate final solution from the Pareto front solution set, thereby generating the most reasonable dynamic control strategy at this moment. This makes the entire system no longer rigidly execute fixed rules, but can intelligently adapt to complex and ever-changing real road conditions and achieve optimal coordinated emission reduction in a global sense.

[0030] The control actions generated by the command execution unit include adjusting the engine ignition angle and injection parameters, correcting the urea injection MAP, triggering the active regeneration of the diesel particulate filter, adjusting the duty cycle signal of the electronic crankcase ventilation valve to change the opening degree, and controlling the opening timing and duration of the fuel evaporation solenoid valve.

[0031] For example, in terms of internal engine control, the command execution unit sends instructions to the engine electronic control unit to adjust the ignition angle and fuel injection parameters. Precise adjustment of the ignition advance angle optimizes the combustion process, thereby affecting the generation of nitrogen oxides and particulate matter. Simultaneously, comprehensive adjustment of the fuel injection quantity, injection timing, and injection pressure directly controls combustion efficiency and fuel economy, thus reducing carbon dioxide emissions. For the coordinated management of the aftertreatment system, the unit corrects the MAP (Modular Mapping) of urea injection in the selective catalytic reduction system. The MAP is essentially a preset lookup table of urea injection quantity based on engine speed and load. Correction means adjusting the specific values ​​in this table in real time according to cloud-based instructions, thereby ensuring nitrogen oxide conversion efficiency while avoiding excessive nitrogen oxide emissions. To address ammonia escape caused by injection, the command execution unit can trigger the active regeneration program of the diesel particulate filter when necessary. This involves increasing the exhaust temperature to burn off the captured particulate matter and restore its filtering capacity. For active intervention in the evaporative emission control system, the command execution unit precisely changes the opening degree of the electronic crankcase ventilation valve by adjusting its duty cycle signal. This controls the flow and direction of blow-by gas in the crankcase, ensuring that the fuel-air mixture is effectively introduced into the intake system for combustion, thus preventing oil vapor from being directly released into the atmosphere. The command execution unit also controls the opening timing and duration of the fuel evaporation solenoid valve, which determines when and at what rate the fuel vapor adsorbed in the carbon canister is desorbed and sent to the engine for combustion. This ensures both the adsorption capacity of the carbon canister and effective control of volatile organic compound emissions.

[0032] In the online diagnostic and fault-tolerant unit, When the internal pressure of the crankcase is continuously higher than the first pressure threshold, it is determined that the crankcase ventilation valve is blocked and a first alarm signal is generated. When the pressure in the fuel vapor carbon canister is continuously lower than the second pressure threshold, it is determined that there is a leak in the evaporation system and a second alarm signal is generated. When the calculated carbon canister saturation value exceeds the capacity threshold, a request is sent to the instruction execution unit to start the carbon canister desorption and regeneration process.

[0033] For example, in diagnosing a clogged crankcase ventilation valve, the crankcase ventilation system's function is to properly guide the exhaust gas that enters the crankcase during engine operation. Under normal operating conditions, the ventilation valve adjusts its opening to maintain the internal pressure within a reasonable range. The online diagnostic and fault-tolerant unit continuously monitors the internal pressure value through a piezoresistive crankcase pressure sensor. The system sets a first pressure threshold, which is a calibrated safety upper limit. When the monitored pressure value continuously exceeds this threshold, the diagnostic logic within the online diagnostic and fault-tolerant unit will determine that the ventilation pipe or valve itself may be clogged with sludge or carbon deposits, preventing exhaust gas from being discharged smoothly and causing an abnormal increase in pressure. At this point, the system will determine that a blockage fault has occurred and immediately generate a first alarm signal to alert the driver or fleet manager that maintenance is required.

[0034] For diagnosing leaks in the fuel evaporation system, which is a closed system designed to collect fuel vapors evaporating from the fuel tank and prevent them from being directly released into the atmosphere, this unit monitors the system pressure via a pressure sensor integrated into the carbon canister. The system has a set second pressure threshold, a calibrated lower limit of vacuum. Under normal circumstances, the engine generates a vacuum during operation, and the system should maintain a certain negative pressure. If the detected pressure remains consistently below this threshold (i.e., the vacuum level drops significantly or even becomes positive), the diagnostic logic determines that the system's seal has been compromised, and a leak is likely present, allowing outside air to enter or fuel vapor to escape—a serious emissions fault. At this point, the system generates a second alarm signal to warn of the potential leak risk.

[0035] For preventative maintenance strategies regarding carbon canister saturation, the adsorption capacity of the fuel vapor carbon canister is limited. The online diagnostic and fault-tolerant unit calculates the amount of fuel vapor adsorbed in the carbon canister, i.e., saturation, in real time based on existing algorithms (such as reference fuel tank level, mileage, temperature, etc.). The system sets a capacity threshold, representing the maximum safe adsorption capacity of the carbon canister. When the calculated saturation exceeds this threshold, it means that the carbon canister is about to be fully loaded, and its adsorption efficiency will decrease, thus failing to effectively capture fuel vapor and facing the risk of exceeding the limit. At this time, the online diagnostic and fault-tolerant unit will not wait for the fault to occur, but will actively send a request command to the command execution unit. The command execution unit will then drive the relevant actuators (such as controlling the opening and closing of solenoid valves) to start the carbon canister desorption and regeneration program, that is, using the vacuum of the engine intake manifold to blow the adsorbed fuel vapor into the engine to burn it off, thereby restoring the adsorption capacity of the carbon canister.

[0036] This embodiment also includes an engine bench calibration database, which stores fuel consumption and emission characteristic maps covering the entire operating condition universal characteristic range obtained through engine bench tests; the collaborative optimization unit calls the data in the database as a basic reference when generating dynamic control strategies, and executes a collaborative strategy to delay or reduce urea injection to suppress ammonia escape under low engine load conditions.

[0037] For example, the engine bench calibration database is a pre-established high-precision data warehouse. The data in it comes from engine bench tests under strict control conditions. These tests measure and record the core performance indicators of the engine at countless different speed and load combinations, forming a universal characteristic map covering the entire operating range. This means that for any current engine operating state, the system can quickly query the database to find its typical fuel consumption and original emission levels under this state in history. When generating dynamic control strategies, the collaborative optimization unit will call the data in this database as a basic reference. For example, when the system senses that the engine is currently at a certain operating point through real-time data, it will immediately refer to the knowledge learned from the database: at this point, what is the original basic nitrogen oxide emission of the engine, what is the particulate matter generation trend, and what is the fuel consumption rate? Thus, the optimization unit can more accurately assess the possible changes and effects of its control commands.

[0038] Characteristics of low-load operation: When the engine is running at low load, its exhaust temperature is usually low. At this time, the catalyst in the downstream selective catalytic reduction system may not be able to reach its most efficient operating temperature window; ammonia escape risk: If urea is still injected in the normal mode under these conditions, the ammonia produced by urea decomposition cannot be fully reduced by the catalyst, and the excess ammonia that has not participated in the reaction will be directly discharged into the atmosphere, causing secondary pollution, which is ammonia escape.

[0039] To address low-load operating conditions and the risk of ammonia slip, collaborative optimization is required. Based on knowledge obtained from the test bench database, the system intelligently selects to delay the start time of urea injection or reduce the injection volume of urea. Although this may slightly sacrifice the instantaneous conversion efficiency of NOx, it successfully avoids more serious ammonia slip problems and achieves collaborative control of the overall emissions of multiple pollutants, reflecting the purpose of multi-objective optimization.

[0040] The regulatory database management unit is used to receive geographic location information sent from the vehicle-mounted global positioning system receiver or the vehicle-to-everything (V2X) communication module, and automatically match and load regulatory limit files that meet the current emission requirements of the region based on the location information, thereby realizing automatic switching of the applicable regulatory areas.

[0041] For example, in order to intelligently address the key characteristics of differences in emission regulations across different countries and regions, The regulatory library management unit can provide the collaborative optimization unit with accurate and real-time updated regulatory boundary constraints. The regulatory library management unit continuously receives geographic location information streams from outside the vehicle. These streams have two main sources: one is the real-time latitude and longitude coordinates provided by the onboard GPS receiver; the other is network location data such as base station positioning obtained through the vehicle network communication module. This ensures that the vehicle can be accurately located in any area with GPS signal or mobile network coverage. The regulatory database management unit has a digital regulatory database. This database not only contains the texts of different regulations such as China's National VI emission standard, Euro VI emission standard, and US EPA standard, but also converts these texts into limit charts and test procedures that the system can directly understand. These charts define in detail the instantaneous and cumulative limits for pollutants such as nitrogen oxides and particulate matter under various test conditions. When the regulatory database management unit receives the vehicle's geographical location information, it immediately initiates a matching query process. For example, if the system identifies that a vehicle is exiting China through the Khorgos border crossing in Xinjiang and entering Kazakhstan, it will automatically query the currently effective emission regulations in Kazakhstan and load the corresponding regulatory limit files from the database. At the same time, it can unload the no longer applicable China VI standard files. After the matching and loading are completed, this new set of regulatory limits applicable to the current region immediately takes effect and becomes a new boundary condition that the collaborative optimization unit must strictly adhere to when performing calculations. All decisions of the optimization unit, whether adjusting engine fuel injection and ignition or controlling urea injection, must ensure that the emission concentration meets the newly loaded regulatory limits as an absolute prerequisite.

[0042] When the online diagnostic and fault-tolerant unit generates the first alarm signal or the second alarm signal, it sends a degraded operation command to the command execution unit. The command execution unit then controls the engine electronic control unit to enter the torque limiting mode and stores the corresponding fault code in the on-board diagnostic system.

[0043] For example, when the online diagnostics and fault tolerance unit determines, based on its monitoring logic, that there is a blockage in the crankcase ventilation valve or a leak in the fuel evaporation system, it will generate a corresponding alarm signal. In this case, the system's response is not limited to recording a fault code or illuminating a warning light on the dashboard. The online diagnostics and fault tolerance unit will immediately issue a clear degraded operation command to the command execution unit. Upon receiving this command, the command execution unit will control the engine's electronic control unit to force it into a pre-set torque limiting mode. In this mode, the engine's electronic control unit will reduce the engine's maximum output torque and power, meaning the driver will feel the vehicle accelerating. The system limits engine power and maximum speed to prevent the engine from continuing to operate under high load in abnormal conditions such as poor ventilation or fuel vapor leakage, which could lead to more serious mechanical damage or safety hazards. It also indirectly controls the combustion process by limiting the engine's operating range, thereby suppressing additional harmful emissions that may be generated due to system failures and ensuring that vehicle emissions levels do not get completely out of control. In addition, the system will follow standard on-board diagnostic protocols to permanently store information such as the type of fault, the time of occurrence, and the degraded measures taken by the system in the form of specific fault codes in the on-board diagnostic system. This provides accurate data for subsequent maintenance and repair. Repair personnel can read these codes with diagnostic instruments to quickly locate and repair faults.

[0044] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An intelligent collaborative emission control system for heavy-duty vehicles, characterized in that, The system includes: The data sensing unit, which is deployed in the vehicle engine and after-treatment system, is used to collect operating status signals including engine speed, load, intake air temperature, exhaust air temperature, crankcase internal pressure, fuel vapor canister pressure and saturation, as well as emission signals such as nitrogen oxide concentration, particulate matter number and volatile organic compound concentration through emission sensors. The data interaction unit is communicatively connected to the data sensing unit and is used to package and upload the encapsulated sensing data to the cloud service platform through the vehicle wireless communication module, and to receive downlink control commands from the cloud service platform. The cloud-based intelligent decision-making module, deployed on a cloud service platform, includes: The emission prediction unit uses a prediction model based on a long short-term memory neural network to perform time-series analysis on the received real-time sensor data in order to predict the concentration change trend of each emission within a specific future time window. The collaborative optimization unit is used to optimize and solve the dynamic control strategy based on the output of the emission prediction unit, combined with the current engine operating point and the working status of the aftertreatment device, with the goal of simultaneously minimizing nitrogen oxide emissions, particulate matter emissions and carbon dioxide emissions. The regulatory library management unit stores limit charts and test procedures for multiple sets of emission regulations, and can automatically select the applicable regulatory version based on vehicle location information or user input, providing boundary constraints for the collaborative optimization unit. The command execution unit is connected to the engine electronic control unit, the selective catalytic reduction system urea injection controller, the diesel particulate filter regeneration controller, the electronic crankcase ventilation valve, and the fuel evaporation solenoid valve. It is used to receive and parse the control commands issued by the cloud-based intelligent decision module and drive the corresponding actuators to make adjustments. The online diagnostic and fault-tolerant unit interacts with the data sensing unit and the instruction execution unit to monitor the working status of the crankcase ventilation system and the fuel evaporation system in real time, perform fault diagnosis based on the comparison of sensor readings with preset thresholds, and activate a predefined adaptive compensation strategy when an anomaly is detected.

2. The system as described in claim 1, characterized in that, The data sensing unit includes a high-precision nitrogen oxide sensor, a particulate matter number counter, a photoionization volatile organic compound sensor, a piezoresistive crankcase pressure sensor, and a fuel vapor canister with integrated pressure and temperature sensing functions.

3. The system as described in claim 1, characterized in that, The input parameter set of the emission prediction unit includes at least engine speed, engine output torque, engine intake manifold temperature, upstream exhaust temperature of the selective catalytic reduction system, and cumulative urea injection quantity; the prediction model is jointly trained by bench tests and actual road emission test data, and can predict whether emissions are at risk of exceeding standards several seconds in advance.

4. The system as described in claim 1, characterized in that, The collaborative optimization unit uses the Pareto front optimization method to solve multi-objective optimization problems and has a built-in dynamic weight allocator. This weight allocator can automatically adjust the weight coefficients of different emission optimization objectives according to the current operating conditions of the vehicle. The operating conditions include at least urban congestion conditions, suburban cruising conditions, and high-speed acceleration conditions.

5. The system as described in claim 1, characterized in that, The control actions generated by the command execution unit include adjusting the engine ignition angle and injection parameters, correcting the urea injection MAP, triggering the active regeneration of the diesel particulate filter, adjusting the duty cycle signal of the electronic crankcase ventilation valve to change the opening degree, and controlling the opening timing and duration of the fuel evaporation solenoid valve.

6. The system as described in claim 1, characterized in that, In the online diagnostic and fault-tolerant unit, When the internal pressure of the crankcase is continuously higher than the first pressure threshold, it is determined that the crankcase ventilation valve is blocked and a first alarm signal is generated. When the pressure in the fuel vapor carbon canister is continuously lower than the second pressure threshold, it is determined that there is a leak in the evaporation system and a second alarm signal is generated. When the calculated carbon canister saturation value exceeds the capacity threshold, a request is sent to the instruction execution unit to start the carbon canister desorption and regeneration process.

7. The system as described in claim 1, characterized in that, It also includes an engine bench calibration database, which stores fuel consumption and emission characteristic maps covering the entire operating condition universal characteristic range obtained through engine bench tests; the collaborative optimization unit calls the data in this database as a basic reference when generating dynamic control strategies, and executes a collaborative strategy to delay or reduce urea injection to suppress ammonia escape under low engine load conditions.

8. The system as described in claim 1, characterized in that, The regulatory database management unit is used to receive geographic location information sent from the vehicle-mounted global positioning system receiver or the vehicle-to-everything (V2X) communication module, and automatically match and load regulatory limit files that meet the current emission requirements of the region based on the location information, thereby realizing automatic switching of the applicable regulatory areas.

9. The system as described in claim 6, characterized in that, When the online diagnostic and fault-tolerant unit generates the first alarm signal or the second alarm signal, it sends a degraded operation command to the command execution unit. The command execution unit then controls the engine electronic control unit to enter the torque limiting mode and stores the corresponding fault code in the on-board diagnostic system.

Images