Cooperative self-adaptive control method and system for pre-combustion chamber parameters, dilution degree and combustion phase and computer readable storage medium

Abstract

The invention discloses a cooperative self-adaptive control method and device for parameters, dilution degree and combustion phase of a pre-combustion chamber. The method comprises the following steps: establishing a parameterized pre-combustion chamber structure set K, defining a working condition state vector and a controllable variable, setting constraints including a combustion phase CA50 window, a combustion duration CA10-90 threshold, cyclic variation COVIMEP and an emission threshold, and constructing an efficiency / stability / emission multi-target cost function; the structural layer selects a structural index k * which meets the constraint and has the minimum cost within K, and the control layer solves the optimal control quantity u * under the k * and enables target output to enter a set window through stacking control of CA50 closed loop + CA10-90 limitation; cA50, CA10-90 and COVIMEP are calculated on line based on the cylinder pressure signals, and parameter online identification and rolling optimization are carried out; when COVIMEP, UHC / NOx or CA10-90 is close to a threshold value, rapid switching between structural layer reevaluation and an equivalent structure composed of a replaceable orifice plate / nozzle assembly / adjustable taper angle piece is triggered, and protection such as inverse integral saturation and rate limitation is assisted. According to the scheme, the combustion phase and duration can be stably controlled under the wide-range working condition, circulation fluctuation and emission are reduced, the dilution limit is expanded, and the heat efficiency is improved.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of combustion control of internal combustion engine, and particularly relates to a method and system for coordinated adaptive control of pre-chamber parameters, dilution degree and combustion phase, and a computer readable storage medium. BACKGROUND

[0002] Under the dual constraints of improving thermal efficiency and reducing emissions, pre-chamber jet ignition can accelerate flame propagation, expand dilution limits and suppress knock by generating high-temperature and high-activity flame in a small-volume cavity and igniting the main combustion chamber through multiple injection holes.

[0003] However, pre-chamber structure parameters (injection hole area / number / orientation / cone angle and volume distribution) and dilution degree (lambda / EGR, lambda is air-fuel ratio, EGR is exhaust gas recirculation rate) have coupled effects on combustion phase and duration; when high compression ratio and ultra-lean combustion are concurrent, the combustion window narrows, the cycle fluctuation increases, and the emission trade-off is complex. Existing solutions mostly use separate parameters and offline calibration, and lack online coordination and wide-range adaptation of structure-dilution-phase, making it difficult to balance efficiency, stability and emissions.

[0004] The present application provides a method for coordinated adaptive control of pre-chamber parameters, dilution degree and combustion phase, particularly regarding how to stably control combustion phase and duration under wide-range operating conditions, reduce cycle fluctuation and emissions, expand dilution limits and improve thermal efficiency. SUMMARY

[0005] The present application aims to solve at least one of the technical problems existing in the prior art. To this end, the present application provides a method for coordinated adaptive control of pre-chamber parameters, dilution degree and combustion phase, which aims to stably control combustion phase and duration under wide-range operating conditions, reduce cycle fluctuation and emissions, expand dilution limits and improve thermal efficiency.

[0006] To achieve the above-mentioned purpose, the technical solution adopted by the present application is as follows: a method for coordinated adaptive control of pre-chamber parameters, dilution degree and combustion phase, comprising the following steps:

[0007] S1: Establish a parameterized pre-chamber structure set K, which is composed of at least one or more of the following: total injection hole area, injection hole number and its spatial distribution, injection hole orientation, jet cone angle, and volume distribution ratio of pre-chamber and main combustion chamber;

[0008] S2: Define the operating condition state vector s and the controllable variable u, set the target output y and the constraint Ω, wherein the target output y at least includes the combustion phase CA50 and the combustion duration CA10-90, and the constraint Ω at least includes the combustion phase CA50 window, the combustion duration CA10-90 threshold, the COV IMEP threshold and U HC / NOx threshold;

[0009] S3: Construct a multi-objective cost function J of normalized deviation weighted sum of efficiency / economy, combustion phasing, combustion duration, cycle-to-cycle variability and emissions, and adaptively adjust the weights based on the load / dilution regime;

[0010] S4: Perform a two-layer collaborative optimization: the structure layer selects a structure index k in the pre-chamber structure set K that satisfies the constraints Ω and minimizes the multi-objective cost function J * ; the control layer solves the optimal control amount u * under the structure index k * and makes the target output y enter the target window with closed-loop control of combustion phasing CA50 and limited combustion duration CA10-90;

[0011] S5: On-line calculate the combustion phasing CA50, combustion duration CA10-90 and COV IMEP based on the cylinder pressure signal, and use recursive least squares and / or Kalman filtering to on-line identify the weight of the multi-objective cost function J according to which to roll-optimize the controllable variables u;

[0012] S6: When the COV IMEP , U HC , NOx or combustion duration CA10-90 approaches the threshold, trigger the structure layer to re-evaluate and fallback, switch to an equivalent structure within the pre-chamber structure set K, and re-solve the optimal control amount u * ; implement anti-integration saturation and rate limiting on the control amount, and set the maximum pressure rise rate and peak pressure phase.

[0013] The pre-chamber structure set K is organized according to one or a combination of the following subsets: total jet hole area grading, jet hole number and spatial distribution serialization, jet hole orientation and jet cone angle combination serialization, pre-chamber and main combustion chamber volume distribution ratio serialization.

[0014] The constraints Ω include: the combustion phasing CA50 is within a preset window after top dead center; the combustion duration CA10-90≤CA10-90 * ; the COV IMEP ≤C * ; the U HC ≤U * ; the NOx≤N * .

[0015] The multi-objective cost function J increases the penalty weight on the COV IMEP and CA10-90 when the dilution approaches the limit.

[0016] The control layer adopts a stacked control of "CA50 closed loop + CA10-90 feedforward / feedback coupling", and links the dilution actuator, ignition system and main / pre-combustion chamber injection process.

[0017] The actuator includes one or a combination of the following: throttle valve, turbocharger, EGR valve, gas dilution device, ignition energy / ignition advance angle adjustment mechanism, and main combustion chamber / pre-combustion chamber injection metering and timing mechanism.

[0018] The online identification employs at least one of recursive least squares and / or Kalman filtering to update the sensitivity matrix of the control input to the target output.

[0019] The equivalent structure is achieved through replaceable nozzle plates, replaceable nozzle assemblies, and / or adjustable jet cone angle components, allowing for rapid switching of structural index k without altering the engine body. * .

[0020] The present invention also provides a cooperative optimization control system for implementing the aforementioned cooperative adaptive control method, comprising a target generation unit, a state estimation unit, a structure selection unit, a continuous optimization unit, an online identification unit, an execution coordination unit, a protection and backoff unit, and a data and knowledge base;

[0021] The structure selection unit is used to select the structure index k from the pre-combustion chamber structure set K that satisfies constraint Ω and minimizes the multi-objective cost function J. * Continuous optimization units are used at structure index k * Solve and issue the optimal control quantity u * The protection and backoff unit is used to trigger structural reassessment and degradation control when the structure exceeds or approaches a threshold.

[0022] The present invention also provides a computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the steps of the cooperative adaptive control method.

[0023] The collaborative adaptive control method of pre-combustion chamber parameters, dilution degree and combustion phase of the present invention achieves stable control of combustion phase and duration under a wide range of operating conditions, reduces cycle fluctuations and emissions, expands the dilution limit and improves thermal efficiency. Attached Figure Description

[0024] This manual includes the following figures, which illustrate the following:

[0025] Figure 1 This is a schematic diagram of the test bench and measurement setup;

[0026] Figure 2 This is a schematic diagram of the pre-combustion chamber structure parameters;

[0027] Figure 3aPlots of ITE and ISFC vs. λ for different CR;

[0028] Figure 3b Plots of CA50 and CA10-90 vs. λ for different CR;

[0029] Figure 4a Plots of ITE and ISFC vs. λ for different injection cone angle;

[0030] Figure 4b Plots of CA50 and CA10-90 vs. λ for different injection cone angle;

[0031] Figure 4c Plots of COV IMEP vs. λ for different injection cone angle;

[0032] Figure 5a Plots of ITE and ISFC vs. λ for different orifice area;

[0033] Figure 5b Plots of CA50 and CA10-90 vs. λ for different orifice area;

[0034] Figure 5c Plots of COV IMEP vs. λ for different orifice area;

[0035] Figure 5d Plots of unburned HC emissions vs. λ for different orifice area;

[0036] Figure 6a Plots of ITE and ISFC vs. λ for different orifice angle;

[0037] Figure 6b Plots of CA50 and CA10-90 vs. λ for different orifice angle;

[0038] Figure 6c Plots of COV IMEP vs. λ for different orifice angle;

[0039] Figure 6d Plots of unburned HC emissions vs. λ for different orifice angle;

[0040] Plots are labeled as:

[0041] 1, optical single cylinder engine; 2, pre-chamber; 3, fuel pump; 4, fuel filter; 5, fuel tank; 6, throttle; 7, intake intercooler; 8, intake control system; 9, air filter; 10, intake; 11, oxygen sensor; 12, exhaust; 13, dynamometer; 14, combustion analyzer; 15, high-speed camera; 16, main control system; 17, jet flame; 18, jet angle; 19, spray; 20, nozzle; 21, spark plug; 22, pre-chamber injection hole angle; 23, pre-chamber injection hole distribution. DETAILED DESCRIPTION

[0042] The specific embodiments of the present application will be further described in the following with reference to the drawings, and by describing the embodiments, in order to help the skilled in the art to have a more complete, accurate and in-depth understanding of the concept and technical solutions of the present application, and to help the implementation thereof.

[0043] In a first aspect, the embodiments of the present application provide a pre-chamber parameter, dilution degree and combustion phase coordinated adaptive control method, comprising the following steps:

[0044] S1: Establish a parameterized pre-chamber structure set K, which is composed of at least one or more of the following: total injection hole area, injection hole number and its spatial distribution, injection hole direction, jet cone angle, pre-chamber and main combustion chamber volume distribution ratio;

[0045] S2: Define the working condition state vector s and the controllable variable u, set the target output y and the constraint Ω, wherein the target output y at least includes the combustion phase CA50 and the combustion duration CA10-90, and the constraint Ω at least includes the combustion phase CA50 window, the combustion duration CA10-90 threshold, the COV IMEP threshold and U HC the unburned hydrocarbon / NOx threshold; the unburned hydrocarbon threshold refers to the maximum allowable concentration of unburned hydrocarbon in engine exhaust; the NOx threshold refers to the maximum allowable concentration of nitrogen oxides in engine exhaust; HC

[0046] S3: Construct a multi-objective cost function J of the normalized deviation weighted sum of efficiency / economy, combustion phase, combustion duration, cycle stability and emission, and adaptively adjust the weight based on the load / dilution interval;

[0047] S4: Perform two-layer coordinated optimization: the structure layer selects a structure index k * in the pre-chamber structure set K that satisfies the constraint Ω and minimizes the multi-objective cost function J * ; the control layer solves the optimal control amount u * in the lower layer based on the structure index k * , and performs a layer-by-layer control of the combustion phase CA50 closed loop and the combustion duration CA10-90 limited to make the target output y enter the target window;

[0048] S5: online calculate combustion phase CA50, combustion duration CA10-90 and COV based on cylinder pressure signal IMEP , online identify using recursive least square and / or Kalman filter Accordingly, the controllable variable u and the weight of the multi-objective cost function J are optimized in rolling;

[0049] S6: when COV IMEP , U HC , NOx or combustion duration CA10-90 approaches threshold, trigger structure layer reevaluation and fallback, switch to equivalent structure within prechamber structure set K, and re-solve optimal control amount u * ; Implement anti-integral saturation and rate limit on control amount, and set maximum pressure rise rate and peak pressure phase.

[0050] Specifically, in the embodiment of the application, to solve the problems of narrow combustion window, increased cycle fluctuation, difficult to stabilize and control phase and duration, and difficult to balance efficiency and emission of prechamber jet ignition (PCJI) internal combustion engine under high compression ratio and ultra-lean parallel conditions, a coordinated adaptive control method of prechamber parameters, dilution degree and combustion phase is proposed, which is applied to a spark-ignition internal combustion engine equipped with prechamber jet ignition (PCJI). Through the two-layer coordination framework of "structure layer discrete selection + control layer continuous optimization", stable constraint and efficiency improvement of key combustion characteristic quantities in a wide operating condition domain are realized.

[0051] In the above step S1, a parameterized description of prechamber structure set K is established, and the elements of K are characterization; Establish operating condition state vector s = [n e , BMEP, T in , p in , CR, …] and controllable variable u = [λ / EGR, ST, inj, …] (where ST is ignition timing, and inj is injection related strategy parameter); Define target output y = [CA50, CA10-90, COV IMEP , NOx, U HC , ITE (indicated thermal efficiency) / ISFC (indicated specific fuel consumption)].

[0052] In the above step S1, the prechamber structure set K is organized according to one or a combination of the following subsets: total nozzle area classification, nozzle number and spatial distribution serialization, nozzle orientation and jet cone angle combination serialization, prechamber and main combustion chamber volume distribution ratio serialization.

[0053] In the above step S1, the total nozzle area (A pc) is the sum of the total area of all injection holes on the pre-chamber wall surface for injecting flame, which determines the intensity and speed of the flame jet. The larger the total area, the greater the flow and the faster the speed of the flame jet, which can ignite the main combustion chamber mixture faster, shorten the combustion duration (CA10-90), and is suitable for high-speed and high-load conditions that require rapid combustion to release energy. The smaller the total area of the injection hole: the slower the jet speed and the smaller the flow, the slower the speed of igniting the main mixture, which is suitable for low-speed and low-load conditions to avoid rapid combustion causing engine vibration. By adjusting the total area of the injection hole, the combustion demand under different dilution rates is adapted.

[0054] In the above step S1, the number of injection holes (n pc ) is the total number of injection holes on the pre-chamber.

[0055] In the above step S1, the injection hole direction refers to the direction of the injection hole, and the jet cone angle refers to the diffusion angle formed after the flame jet is injected from the injection hole. The injection hole direction / jet cone angle determines the target area of the flame jet, ensuring that the flame can accurately cover the mixture enrichment area of the main combustion chamber.

[0056] In the above step S1, the volume distribution ratio (V pc / V mc ) of the pre-chamber to the main combustion chamber is the ratio of the internal volume (V pc ) of the pre-chamber to the internal volume (V mc ) of the main combustion chamber, which determines the concentration of ignition energy inside the pre-chamber.

[0057] In the above step S1, the above parameters are combined into a pre-chamber structure set K, the core of which is to adapt to the wide range of engine operating conditions - the demand for pre-chamber structure is completely different under different conditions:

[0058] High-speed and high-load conditions: require a structure with large total injection hole area, moderate cone angle, and slightly larger volume ratio to ensure fast and strong ignition;

[0059] Low-speed and low-load conditions: require a structure with small total injection hole area, uniform distribution, and slightly smaller volume ratio to ensure stable combustion and no vibration;

[0060] Super-dilution and high EGR conditions: require a structure with large cone angle and large volume ratio to ensure wide flame coverage and sufficient energy to avoid misfire.

[0061] In the above step S2, the target interval and hard constraint for high efficiency and stability are set:

[0062] Target combustion phase: CA50∈[CA50 min ,CA50max ] (e.g. around the top dead center after-effect range window);

[0063] Target combustion duration: CA10-90 no more than threshold CA10-90 * ;

[0064] Stability and emissions: COV IMEP ≤C * , U HC ≤U * , NOx≤N * ;

[0065] Feasible region: subject to λ max (or EGR upper limit), intake / cooling temperature, compression ratio CR and mechanical / thermal load boundaries.

[0066] In the above step S2, the constraints Ω include: the combustion phase CA50 is within a preset window after the top dead center; the combustion duration CA10-90 ≤ CA10-90 * ; COV IMEP ≤C * ; U HC ≤U * ; NOx≤N * .

[0067] CA10 refers to the crank angle corresponding to 10% of the total heat release of the fuel during the combustion process; CA90 refers to the crank angle corresponding to 90% of the total heat release of the fuel during the combustion process; CA10-90 is the numerical difference between CA90 and CA10, which represents the angle through which the engine crankshaft rotates from the beginning of effective combustion to the end of effective combustion. The smaller the angle, the faster the combustion speed; the larger the angle, the slower the combustion speed.

[0068] In the above step S3, a multi-objective cost function J is constructed by linearly combining the normalized deviations of the indicators of efficiency / economy, combustion phase, combustion duration, cycle stability and emissions according to weights; the weights are adaptively adjusted according to the load and dilution interval, and the penalty weights of COV IMEP and CA10-90 are increased when the dilution approaches the limit, to prioritize ignitability and propagability.

[0069] Multi-objective cost function J:

[0070] In the above step S3, the multi-objective cost function J increases the penalty weights of COV IMEP and CA10-90 when the dilution approaches the limit.

[0071] In the above step S3, the normalized deviation of efficiency / economy: taking the ideal highest indicated thermal efficiency ITE0 (or the ideal lowest indicated specific fuel consumption ISFC0, inversely related) as the reference, the actual value is ITE (or ISFC). If taking the indicated thermal efficiency ITE as the efficiency index, its normalized deviation can be expressed as: δ ITE = 1 - (ITE / ITE0), where ITE0 is the reference or ideal indicated thermal efficiency. The closer the actual ITE is to ITE0, the closer δ ITE is to 0; when the actual ITE is lower than ITE0, δ ITE is positive, intuitively reflecting the degree of efficiency loss.

[0072] The normalized deviation of combustion phase (CA50): taking the CA50 target window as the reference, if the actual CA50 is within the window, the deviation = 0.

[0073] The normalized deviation of combustion duration (CA10-90): taking the CA10-90 threshold CA10-90 * as the reference, the actual value is CA10-90. If the actual CA10-90 ≤ CA10-90 * , the deviation = 0.

[0074] The normalized deviation of cycle stability (COV IMEP ): taking the COV IMEP threshold C * as the reference, the actual value is COV IMEP . If the actual COV IMEP ≤ C * , the deviation = 0.

[0075] The normalized deviation of emissions (U HC / NOx): taking the U HC threshold U * , NOx threshold N * as the reference, the deviation of the two types of emissions is calculated respectively. Taking NOx as an example, if the actual NOx ≤ N * , the deviation = 0

[0076] In the above step S4, two-layer collaborative optimization is performed. Structure layer (discrete): selecting the structure index k * in K that satisfies the constraints and makes J attainable minimum (facing wide-range adaptive selection for different speeds / loads and environmental temperatures / pressures);

[0077] Control layer (continuous): under the given k * , solving min u J(u, k * |s) to obtain u *= [λ / EGR,ST,inj,...]; closed-loop correction for CA50, feed-forward-feedback coupled correction for CA10-90.

[0078] In step S4 above, the control layer adopts the layering control of "CA50 closed loop + CA10-90 feed-forward / feedback coupling", and links the dilution execution, the ignition system and the main / pre-chamber injection process.

[0079] The execution includes one or a combination of the following: throttle, supercharger, EGR valve, fuel gas dilution device, ignition energy / ignition advance angle adjusting mechanism, main chamber / pre-chamber injection metering and timing mechanism.

[0080] In step S5 above, online identification and self-learning are performed:

[0081] Based on the cylinder pressure signal, CA50, CA10-90 and COV are calculated in real time IMEP , and the recursive least squares / Kalman method is used to update the sensitivity online ; the observation data are written into the structure- working condition library for subsequent rolling optimization and weight self-tuning.

[0082] In step S5 above, the rolling optimization is based on the sensitivity matrix identified in real time, and the controllable variables u and the weight of the multi-objective cost function J are optimized once every control period, to ensure that the control strategy always adapts to the current working condition.

[0083] In step S6 above, u * is distributed to the dilution execution (throttle / supercharger / EGR / fuel gas dilution), the ignition system (ignition advance angle / energy), the injection system (injection time and distribution ratio of the main chamber and the pre-chamber, if applicable), etc., to maintain the target CA50 and the controlled combustion duration; for transient working conditions, a conservative fallback strategy is used, which prioritizes stability and emission red lines.

[0084] The equivalent structure is realized by replaceable nozzle plate, replaceable nozzle assembly and / or adjustable jet cone angle component, to quickly switch the structure index k * without changing the engine body.

[0085] When λ or environmental disturbance causes COV to approach the limit, NOx / U HC to reach the top or CA10-90 to exceed the threshold, the structure layer re-evaluation is triggered; if there is a better structure within K that can make J decrease and meet the thermal-mechanical constraints, the structure switching (or selection of equivalent "equivalent structure" - realized by replaceable nozzle plate / nozzle assembly / adjustable cone angle component) is performed, and u * is solved again.

[0086] The system employs hard limiting and soft degradation logic for cylinder pressure rise rate, exhaust temperature, and misfire / pre-ignition criteria; and uses inverse integral saturation and rate limiting for u to avoid secondary oscillations caused by actuator saturation.

[0087] Secondly, embodiments of the present invention also provide a cooperative optimization control system for implementing the cooperative adaptive control method of the above embodiments, comprising:

[0088] Target generation unit: Generates y based on efficiency / emissions / stability strategies * With weight ω i ;

[0089] State estimation unit: Integrates cylinder pressure / torque / intake / cooling / emission information to estimate s and the feasible region;

[0090] Structure selection unit: Calculate arg min k∈K J and output k * ;

[0091] Continuous optimization unit: solving arg min k∈K J(·) and output u * ;

[0092] Online identification unit: updating the model and sensitivity matrix;

[0093] Execution Coordination Unit: will u * Distribute to actuators such as dilution, ignition, and injection;

[0094] Protection and rollback unit: anomaly detection, threshold limiting, event-triggered switching, and safe rollback;

[0095] Data and Knowledge Base: Records <s,k,u,y> are used for rolling learning and recalibration.

[0096] The structure selection unit is used to select the structure index k from the pre-combustion chamber structure set K that satisfies constraint Ω and minimizes the multi-objective cost function J. * Continuous optimization units are used at structure index k * Solve and issue the optimal control quantity u * The protection and backoff unit is used to trigger structural reassessment and degradation control when the structure exceeds or approaches a threshold.

[0097] Thirdly, embodiments of the present invention also provide a computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the steps of the cooperative adaptive control method described above.

[0098] The aforementioned collaborative adaptive control method and system for pre-combustion chamber parameters, dilution degree, and combustion phase has the following advantages:

[0099] 1) Wide-range feasibility: Through the two-layer collaborative optimization of "structure layer + control layer", the stable constraints of CA50 and CA10-90 in the high CR x ultra-lean combustion interval are realized, and the COV is obviously reduced IMEP and misfire probability;

[0100] 2) Efficiency-emission balance: Under the premise of meeting stable combustion, the ITE / ISFC is taken as the benefit item, and the NOx / U HC is taken as the constraint item for joint optimization to avoid the side effects caused by single-index extremum;

[0101] 3) Self-adaptation and portability: The online identification and rolling learning mechanism reduces the dependence on artificial repeated calibration and is applicable to different compression ratios, different pre-chamber templates, and different dilution paths;

[0102] 4) Safety robustness: The event-triggered structure re-evaluation and fallback strategy ensures controllability and recoverability under environmental disturbance and actuator saturation.

[0103] Embodiment 1: Apparatus and object-parameterized structure set K, measurement and target setting

[0104] 1) Test / control object and measurement chain

[0105] The object is a spark-ignition single-cylinder thermodynamic engine equipped with a pre-chamber jet ignition (PCJI), which can quickly replace the pre-chamber nozzle plate / nozzle assembly (to realize "equivalent structure" switching), and the cylinder pressure sampling is used to calculate CA50, CA10-90 and COV IMEP , λ / emission / charge and cooling boundary are controlled and collected. The engine allows systematic comparison of pre-chamber geometries under different compression ratios (CR) and dilution ratios (λ / EGR).

[0106] 2) Pre-chamber structure set K and state / control variable definition

[0107] The pre-chamber structure set K is characterized by the parameter group {total nozzle area A pc , nozzle number n pc and distribution, hole direction / jet cone angle θ j , pre-chamber-main chamber volume distribution ratio V pc / V mc}; the working condition state vector s = [n e , BMEP, T in , p in , CR, …], the controllable variable u = [λ / EGR, ST (ignition timing), inj (injection-related parameters), …]; and the output y = [CA50, CA10-90, COV IMEP , NOx, U HC , ITE / ISFC].

[0108] 3) Objective and constraint setting

[0109] Objective: high efficiency and stability: set CA50 in small window after TDC, CA10-90 not exceeding threshold, COV IMEP not higher than threshold, U HC / NOx not exceeding limit; feasible region affected by max / EGR upper limit, thermal / mechanical boundary and CR.

[0110] 4) Multi-objective cost function and weight self-tuning

[0111] Construct normalized deviation weighted sum of efficiency / phase / duration / stability / emission, dynamically increase penalty weight of COV and CA10-90 item when in different load or close to dilution limit, prioritize ignitability / propagability and phase accessibility.

[0112] Embodiment 2: Two-layer collaborative optimization (structure layer + control layer) and closed-loop strategy

[0113] 1) Structure layer (discrete optimization)

[0114] Determine accessibility and constraint feasibility within K, select structure index k that minimizes cost function * (e.g. filter in subgroups such as total area of injection holes, jet cone angle, hole direction distribution, etc.).

[0115] 2) Control layer (continuous optimization)

[0116] Given k * , solve u * = [λ / EGR, ST, inj, …]; adopt "CA50 closed loop + CA10-90 limited" layered control: CA50 derived from cylinder pressure as main loop, CA10-90 suppressed by fast / slow combustion through feedforward / feedback coupling and rate limitation; if necessary, link dilution actuators, ignition energy and main / pre-chamber injection history.

[0117] 3) Online identification and rolling optimization

[0118] Based on real-time calculation of CA50, CA10-90 and COV IMEP from cylinder pressure, recursively least square / Kalman filter online identification sensitivity, and rolling correction of optimal solution of u and each item weight.

[0119] 4) Protection and fallback

[0120] Apply anti-integral saturation and rate limitation to control variables; set maximum pressure rise rate, peak pressure phase, misfire / pre-ignition protection thresholds, prioritize stability and emission boundary when triggered.

[0121] Example 3: Typical procedure under steady-state conditions (corresponding to Figures 3-5)

[0122] 1) Initialization and pre-selection of structures

[0123] Based on the λ—performance / phase / duration relationship of Figures 3a to 5d , two sets of pre-chamber templates are selected into the candidate set K (e.g. nozzle total area 6.8mm 2 and 4.7mm 2 , or jet cone angle 120° and 140° combined as controls). The structure layer is selected with the minimum cost criterion of optimal ITE / ISFC and lowest COV IMEP under the premise of meeting the CA50 window and the upper limit of CA10-90. * .

[0124] 2) Closed-loop optimization and convergence

[0125] With CA50→target window as the main ring, λ / EGR and ST are linked; if CA10-90 is too long, gradually increase the premixing ratio or tighten the jet cone angle; if CA10-90 is too short and is too high, reduce the local reaction activity (slightly reduce the total area of the injection hole or relax the cone angle) and limit the ignition energy until CA10-90 and COV IMEP fall within the threshold.

[0126] 3) Results recording

[0127] Store <s, k * , u * , y> for subsequent rolling optimization and cross-condition migration to provide priori.

[0128] Note: When high CR×high λ is promoted in parallel, the optimal efficiency point moves to the right (λ increases) with the increase of CR, which is used as the reference baseline for target setting and weight adjustment.

[0129] Example 4: Rolling optimization under transient disturbance (step dilution and load variation)

[0130] 1) Disturbance injection

[0131] Apply a λ (or EGR) step to the steady-state solution u * ; if the observed COV IMEP approaches the threshold or CA10-90 exceeds the limit, increase the stability / duration weight and relax the efficiency weight.

[0132] 2) Online re-identification and re-optimization

[0133] The sensitivity matrix is recursively estimated using recent N-cycle data, and the u is quickly corrected by quadratic programming; the rate limit is applied to the actuator to suppress secondary oscillation; after the CA50 / CA10-90 window is met, the efficiency weight is gradually increased to restore the original solution dominated by the income term.

[0134] Embodiment 5: Structure switching and "equivalent structure" fast fallback

[0135] Triggering conditions and strategies

[0136] When the COV IMEP , U HC / NOx or CA10-90 approaches the threshold and the control layer is not reachable, the structure layer re-evaluation is triggered:

[0137] (i) First fine-tune the u steady voltage near k * ; if it is still not reachable,

[0138] (ii) Switch to the equivalent "equivalent structure" (achieved by replaceable orifice plate / nozzle assembly / adjustable cone angle device) within K, and then solve u * to restore CA50 and CA10-90 to the target window.

[0139] The present application has been described above with reference to specific embodiments. Obviously, the specific implementation of the present application is not limited to the above-described manner. As long as various non-essential improvements are made using the method concept and technical solutions of the present application; or without improvement, the above-mentioned concept and technical solutions of the present application are directly applied to other occasions, which are within the protection scope of the present application.

Claims

1. A method of synergistic adaptive control of pre-chamber parameters, dilution level and combustion phasing, characterized in that, The method comprises the following steps: S1: establishing a parameterized pre-chamber structure set K, which is composed of at least one or more of the following: total jet hole area, number of jet holes and its spatial distribution, jet hole orientation, jet cone angle, and volume distribution ratio of pre-chamber to main combustion chamber; S2: define the operating condition state vector s and the controllable variables u, set the target outputs y and the constraints Ω, the target outputs y including at least the combustion phase CA50 and the combustion duration CA10-90, the constraints Ω including at least the combustion phase CA50 window, the combustion duration CA10-90 threshold, the COV IMEP thresholds and U HC NOx thresholds; S3: constructing a multi-objective cost function J of normalized deviation weighted sum of efficiency / economy, combustion phase, combustion duration, cycle stability and emissions, and adaptively adjusting the weights based on the load / dilution interval; S4: Perform two-layer collaborative optimization: structure layer selects structure index k in pre-chamber structure set K that satisfies constraint Ω and minimizes multi-objective cost function J * ; control layer solves optimal control amount u under structure index k * * and makes target output y enter target window with closed-loop of combustion phase CA50 and limited combustion duration CA10-90​ S5: online calculate combustion phase CA50, combustion duration CA10-90 and COV based on cylinder pressure signal IMEP , online identify the model parameters using recursive least square and / or Kalman filter Accordingly, the controllable variable u and the weight of the multi-objective cost function J are rolling optimized. S6: Trigger structure layer reassessment and fallback when COV IMEP , U HC , NOx or combustion duration CA10-90 approaches threshold, switch to equivalent structure within prechamber structure set K and re-solve optimal control u * ; implement anti-integral saturation and rate limiting on control, and set maximum pressure rise rate and peak pressure phase.

2. The method of claim 1, wherein, The pre-chamber structure set K is organized according to one or a combination of the following subsets: total jet hole area classification, number of jet holes and spatial distribution series, jet hole orientation and jet cone angle combination series, and pre-chamber to main combustion chamber volume distribution ratio series.

3. The method of claim 1, wherein, The constraint Ω includes: combustion phase CA50 is located in a preset window after top dead center; combustion duration CA10-90 ≤ CA10-90 * ; COV IMEP ≤ C * ; U HC ≤ U * ; NOx ≤ N * .

4. The method of co-adaptive control according to any one of claims 1 to 3, characterized in that, The multi-objective cost function J increases the penalty weight of COV IMEP with the penalty weight of CA10-90.

5. The method of claim 1, 2 or 3, wherein, The control layer adopts a "CA50 closed loop + CA10-90 feedforward / feedback coupling" layered control, and links the dilution execution device, the ignition system and the main / pre-chamber injection history.

6. The co-adaptive control method of claim 5, wherein, The execution device includes one or a combination of the following: throttle, supercharger, EGR valve, fuel gas dilution device, ignition energy / ignition advance angle adjusting mechanism, main combustion chamber / pre-chamber injection metering and timing mechanism.

7. The method of co-adaptive control according to any of claims 1 to 3, characterized in that, The online identification uses at least one of recursive least squares and / or Kalman filtering to update the sensitivity matrix of the control quantity to the target output.

8. The method of co-adaptive control according to any one of claims 1 to 3, characterized in that, The equivalent structures are achieved by means of replaceable orifice plates, replaceable nozzle assemblies and / or adjustable jet cone angle components to quickly switch the structure index k without changing the engine body * .

9. A synergic optimal control system for implementing the synergic adaptive control method according to any one of claims 1 to 8, characterized in that, It includes a target generation unit, a state estimation unit, a structure selection unit, a continuous optimization unit, an online identification unit, an execution coordination unit, a protection and fallback unit, and a data and knowledge base; wherein the structure selection unit is configured to select a structure index k in the pre-chamber structure set K that satisfies the constraint Ω and minimizes the multi-objective cost function J * , the continuous optimization unit is configured to solve and issue the optimal control quantity u at the structure index k * * The protection and fallback unit is configured to trigger structure reevaluation and degradation control when the threshold is crossed or approached.​ 10. A computer-readable storage medium, characterized in that, A program is stored thereon, which, when executed by a processor, implements the steps of the collaborative adaptive control method of any one of claims 1-8.

Images