FORM – 2
THE PATENTS ACT, 1970
(39 of 1970)
&
THE PATENTS RULES, 2003
COMPLETE SPECIFICATION
(See section 10 and rule 13)
A PREDICTIVE BACKFIRE PREVENTION SYSTEM FOR A HYDROGEN-FUELED
INTERNAL COMBUSTION ENGINE
Applicant(s):
VE COMMERCIAL VEHICLES LTD
102, INDUSTRIAL AREA 1, PITHAMPUR- 454775, DIST. DHAR, MP, INDIA
Inventors:
(1) Abhishek Tripathi
(2) Rahul Dubey
(3) Hardik Lakhlani
(4) Hemant Rathi
(5) Akhilesh Shukla
(6) Sachin Agarwal
The following specification particularly describes the invention and the manner in which it
is to be performed.
2
A PREDICTIVE BACKFIRE PREVENTION SYSTEM FOR A
HYDROGEN-FUELED INTERNAL COMBUSTION ENGINE
FIELD OF THE DISCLOSURE
[0001] This invention generally relates to a field of a safety system for vehicles, and
in particular, to a predictive backfire prevention system for a hydrogen-fueled5
internal combustion engine and method thereof.
BACKGROUND
[0002] The subject matter discussed in the background section should not be
assumed to be prior art merely as a result of its mention in the background section.
Similarly, a problem mentioned in the background section or associated with the10
subject matter of the background section should not be assumed to have been
previously recognized in the prior art. The subject matter in the background section
merely represents different approaches, which in and of themselves may also
correspond to implementations of the claimed technology.
[0003] Hydrogen internal combustion engines are increasingly being developed as15
an alternative propulsion technology due to their ability to reduce carbon-based
emissions while leveraging existing internal combustion engine platforms. In such
engines, hydrogen fuel is introduced into an intake port or combustion chamber and
ignited under conditions that differ significantly from those associated with
conventional hydrocarbon fuels.20
[0004] Hydrogen possesses a very low ignition energy and a high flame propagation
speed when compared to gasoline or diesel fuels. As a result, hydrogen fuel is highly
susceptible to premature ignition in the presence of hot residual exhaust gases,
heated intake valves, spark plug components, or localized hot spots within the intake
system.25
[0005] Premature ignition of hydrogen fuel, commonly referred to as intake
backfire, may occur within an intake manifold or intake port prior to complete
closure of an intake valve. Such backfire events can propagate rapidly in a reverse
direction along the intake flow path due to the fast flame speed of hydrogen.
3
[0006] Intake backfire events are undesirable as they can produce loud acoustic
disturbances, cause transient or sustained loss of engine torque, and, in severe cases,
result in mechanical damage to intake components, sensors, or associated ducting.
Repeated backfire events further degrade drivability and reduce overall engine
reliability.5
[0007] Conventional approaches for mitigating backfire in hydrogen internal
combustion engines are largely reactive in nature. Such approaches typically rely
on detection of backfire after its occurrence, for example by sensing a pressure
spike, temperature increase, or abnormal engine speed fluctuation within the intake
system, at which point corrective measures are applied to recover engine operation.10
[0008] Because reactive backfire detection occurs after ignition has already taken
place, such methods are inherently limited in their ability to prevent damage or
combustion instability during the affected engine cycle. Consequently, these
approaches primarily serve to mitigate the effects of backfire rather than to prevent
its occurrence.15
[0009] In addition to reactive strategies, existing hydrogen engine control systems
frequently employ static or conservative calibration techniques to reduce backfire
risk. Such techniques include fixed late fuel injection timing, reduced fueling rates,
or permanently retarded ignition timing, which are applied across broad operating
conditions to maintain a safety margin.20
[0010] While conservative calibration reduces the likelihood of backfire, it imposes
a persistent compromise on engine performance, combustion efficiency, and
transient response. These static strategies fail to account for dynamic variations in
engine load, speed, temperature, and residual gas conditions, leading to suboptimal
operation under many normal driving scenarios.25
[0011] Another the patent application, “CN102322337B,” titled “Hydrogen fuel
internal combustion engine flashback prevention and abnormal combustion
suppression method,” describes a method for preventing backfire of a hydrogen fuel
internal combustion engine and suppressing abnormal combustion. The prevention
method uses at least two hydrogen injectors to inject hydrogen, and the hydrogen30
injectors adopt a parallel and segmented working mode to inject hydrogen; After
4
the gas valve is opened, the hydrogen injection is delayed for a certain period of
time; the end time of the hydrogen injection is fixed at the moment when the intake
valve starts to close; Inject hydrogen into the valve overlap angle at all times, and
minimize the residence time of hydrogen in the intake pipe. It can also cool the hot
spots inside the intake manifold through the early intake air, reducing the probability5
of the mixture being ignited to reduce the probability of flashback; fixed The
hydrogen injection method with the longest hydrogen injection duration of each
channel, regardless of the load, the starting moment of each hydrogen injection will
not be close to the overlapping angle of the front end of the inlet valve opening;
effectively prevent the occurrence of backfire; and adjust the ignition The10
combination of timing can effectively suppress abnormal combustion.
[0012] Another patent application, "US11591953B2," titled " Method for
controlling hydrogen combustion in a hydrogen internal combusting engine,"
describes A method for controlling hydrogen combustion in a hydrogen internal
combustion engine system includes a combustion chamber linked to an intake port15
via an intake valve, the hydrogen internal combustion engine system comprising a
piston slidably moving between a top dead center position and a bottom dead center
position, characterized by the steps of: injecting water in liquid phase in the intake
port when the piston is between 0 and 40 crank angle degrees before opening of the
intake valve, injecting hydrogen after opening of the intake valve and when the20
piston is between 0 and 60 crank angle degrees after the top dead center position,
stopping hydrogen injection when the piston is between 0 and 100 crank angle
degrees before the bottom dead center position.
[0013] In conventional hydrogen internal combustion engine control systems,
backfire mitigation is typically implemented using reactive detection strategies or25
conservative static calibration maps applied across broad operating conditions.
However, such approaches present several limitations in ensuring both combustion
safety and optimal engine performance. In particular, existing backfire control
techniques are ineffective at predicting imminent backfire conditions prior to fuel
ignition due to the highly reactive nature and low ignition energy of hydrogen fuel.30
Additionally, reliance on delayed detection mechanisms or permanently retarded
5
fuel injection and ignition timing results in corrective actions being applied only
after a backfire event has occurred or in a manner that unnecessarily restricts engine
operation. As a result, conventional hydrogen engine control architectures are prone
to compromised performance, reduced combustion efficiency, transient instability,
and increased susceptibility to intake system stress, especially during high-load,5
low-speed, cold-start, or transient operating conditions.
OBJECTIVES OF THE INVENTION
[0014] The objective of the present invention is to provide a predictive backfire
prevention system for hydrogen-fueled internal combustion engine that proactively
assesses backfire risk prior to fuel ignition.10
[0015] The objective of the present invention is to overcome limitations of reactive
and static backfire mitigation strategies by enabling real-time, cycle-resolved
evaluation of combustion and intake conditions associated with hydrogen backfire.
[0016] The objective of the present invention is to introduce a unified Predictive
Backfire Risk Index (PBRI) that consolidates multiple combustion-related15
parameters into a single normalized metric representing instantaneous backfire risk.
[0017] The objective of the present invention is to integrate predictive analytics and
adaptive learning logic directly within an engine control unit to transform
conventional engine control from reactive operation to preventive combustion
management.20
[0018] The objective of the present invention is to enable real-time sensor fusion of
intake temperature, intake pressure, equivalence ratio, engine speed, crank angle,
and combustion feedback signals for predictive backfire assessment.
[0019] The objective of the present invention is to dynamically adjust engine
control actions, including hydrogen fuel injection timing, injection duration,25
ignition timing, and airflow management, based on the computed PBRI level.
[0020] The objective of the present invention is to provide a coordinated, graduated
preventive control strategy in which corrective actions are proportionally applied
according to backfire risk severity rather than executing abrupt shutdowns.
[0021] The objective of the present invention is to implement an adaptive learning30
mechanism that refines backfire risk estimation using field operating data, thereby
6
accommodating engine aging, fuel variation, and environmental changes without
manual recalibration.
[0022] The objective of the present invention is to achieve predictive backfire
prevention using existing engine sensors and hardware infrastructure, thereby
minimizing system cost, complexity, and packaging requirements.5
[0023] The objective of the present invention is to enhance operational safety,
combustion stability, and engine performance in hydrogen internal combustion
engines by preventing intake backfire events while maintaining optimal efficiency
across varying operating conditions.
SUMMARY10
[0024] The present invention relates to a predictive backfire prevention system for
a hydrogen-fueled internal combustion engine.
[0025] According to an aspect, a predictive backfire prevention system for a
hydrogen-fueled internal combustion engine is disclosed. The predictive backfire
prevention system comprises a plurality of sensors configured to generate real-time15
combustion-related data including at least intake manifold pressure, intake manifold
temperature, engine speed, and hydrogen fuel delivery information. The predictive
backfire prevention system further comprises a processing unit communicatively
coupled to the plurality of sensors and configured to compute, for each combustion
cycle, a set of normalized feature values derived from the real-time combustion-20
related data, the set of normalized feature values including at least an estimate of
residual gas temperature, a local hydrogen equivalence ratio, and a manifold
pressure rate-of-change, compute a raw backfire risk score using a weighted
combination of the set of normalized feature values, map the raw backfire risk score
to a predictive backfire risk index representing a probability of an intake backfire25
event, compare the predictive backfire risk index to one or more predefined
thresholds. The predictive backfire prevention system further comprises a control
interface configured to, when the predictive backfire risk index exceeds at least one
of the predefined thresholds, command at least one corrective engine control action
selected from hydrogen fuel injection modification, ignition timing adjustment, or30
temporary hydrogen fuel cut-off. Further, the predictive backfire risk index is
7
computed prior to occurrence of a backfire event and the corrective engine control
action is applied proactively to suppress intake backfire.
[0026] According to another aspect, a method for operating a predictive backfire
prevention system for a hydrogen-fueled internal combustion engine is disclosed.
The method comprises steps of generating, via a plurality of sensors, real-time5
combustion-related data including at least intake manifold pressure, intake manifold
temperature, engine speed, and hydrogen fuel delivery information. The method
comprises further steps of computing, via a processing unit communicatively
coupled to the plurality of sensors, for each combustion cycle, a set of normalized
feature values derived from the real-time combustion-related data, the set of10
normalized feature values including at least an estimate of residual gas temperature,
a local hydrogen equivalence ratio, and a manifold pressure rate-of-change. The
method comprises further steps of computing, via the processing unit, a raw backfire
risk score using a weighted combination of the set of normalized feature values. The
method comprises further steps of mapping, via the processing unit, the raw backfire15
risk score to a predictive backfire risk index representing a probability of an intake
backfire event. The method comprises further steps of comparing, via the processing
unit, the predictive backfire risk index to one or more predefined thresholds. The
method comprises further steps of commanding, via a control interface, when the
predictive backfire risk index exceeds at least one of the predefined thresholds, at20
least one corrective engine control action selected from hydrogen fuel injection
modification, ignition timing adjustment, or temporary hydrogen fuel cut-off.
Further, the predictive backfire risk index is computed prior to occurrence of a
backfire event and the corrective engine control action is applied proactively to
suppress intake backfire.25
BRIEF DESCRIPTION OF THE DRAWINGS
[0027] The accompanying drawings illustrate various embodiments of systems,
methods, and embodiments of various other aspects of the disclosure. Any person
with ordinary skills in the art will appreciate that the illustrated element boundaries
(e.g., boxes, groups of boxes, or other shapes) in the figures represent one example30
of the boundaries. It may be that in some examples one element may be designed as
8
multiple elements or that multiple elements may be designed as one element. In
some examples, an element shown as an internal component of one element may be
implemented as an external component in another, and vice versa. Furthermore,
elements may not be drawn to scale. Non-limiting and non-exhaustive descriptions
are described with reference to the following drawings. The components in the5
figures are not necessarily to scale, emphasis instead being placed upon illustrating
principles.
[0028] FIG. 1 illustrates a block diagram of a predictive backfire prevention system
for a hydrogen-fueled internal combustion engine, according to an embodiment of
the present disclosure; and10
[0029] FIG. 2 illustrates a flowchart showing a method for operating a predictive
backfire prevention system, according to an embodiment of the present disclosure.
DETAILED DESCRIPTION
[0030] Some embodiments of this disclosure, illustrating all its features, will now
be discussed in detail. The words “comprising,” “having,” “containing,” and15
“including,” and other forms thereof, are intended to be equivalent in meaning and
be open ended in that an item or items following any one of these words is not meant
to be an exhaustive listing of such item or items or meant to be limited to only the
listed item or items. It must also be noted that as used herein and in the appended
claims, the singular forms “a,” “an,” and “the” include plural references unless the20
context clearly dictates otherwise.
[0031] Although any systems and methods similar or equivalent to those described
herein can be used in the practice or testing of embodiments of the present
disclosure, the preferred, systems and methods are now described. Embodiments of
the present disclosure will be described more fully hereinafter with reference to the25
accompanying drawings in which like numerals represent like elements throughout
the several figures, and in which example embodiments are shown. Embodiments
of the claims may, however, be embodied in many different forms and should not
be construed as limited to the embodiments set forth herein. The examples set forth
herein are non-limiting examples and are merely examples among other possible30
examples.
9
[0032] The present invention discloses various embodiments of a predictive
backfire prevention system for a hydrogen-fueled internal combustion.
Embodiments of the present invention comprises a plurality of sensors configured
to generate real-time combustion-related data including at least intake manifold
pressure, intake manifold temperature, engine speed, and hydrogen fuel delivery5
information. Embodiments of the present invention further comprises a processing
unit communicatively coupled to the plurality of sensors and configured to compute,
for each combustion cycle, a set of normalized feature values derived from the real-
time combustion-related data, the set of normalized feature values including at least
an estimate of residual gas temperature, a local hydrogen equivalence ratio, and a10
manifold pressure rate-of-change, compute a raw backfire risk score using a
weighted combination of the set of normalized feature values, map the raw backfire
risk score to a predictive backfire risk index representing a probability of an intake
backfire event, compare the predictive backfire risk index to one or more predefined
thresholds. Embodiments of the present invention further comprises a control15
interface configured to, when the predictive backfire risk index exceeds at least one
of the predefined thresholds, command at least one corrective engine control action
selected from hydrogen fuel injection modification, ignition timing adjustment, or
temporary hydrogen fuel cut-off. Further, the predictive backfire risk index is
computed prior to occurrence of a backfire event and the corrective engine control20
action is applied proactively to suppress intake backfire.
[0033] FIG. 1 illustrates a block diagram of a predictive backfire prevention system
(100) for a hydrogen-fueled internal combustion engine, according to an
embodiment of the present disclosure.
[0034] In some embodiments, the predictive backfire prevention system (100)25
comprises a plurality of sensors (102) operatively associated with a hydrogen-fueled
internal combustion engine. The plurality of sensors (102) are configured to
generate real-time combustion-related data during engine operation. Such data
includes intake manifold pressure, intake manifold temperature, engine speed, and
hydrogen fuel delivery information. The plurality of sensors (102) operate30
continuously across varying engine loads and speeds. The sensor data represents
10
instantaneous intake and combustion conditions. The plurality of sensors (102)
collectively enable real-time observation of parameters relevant to backfire
occurrence. The predictive backfire prevention system (100) utilizes existing engine
sensing infrastructure.
[0035] The plurality of sensors (102) may further include one or more additional5
sensing elements disposed at selected engine locations. In some embodiments, the
plurality of sensors (102) comprises at least one of an ionization sensing circuit, a
knock sensor, a crank-angle position sensor, or a hydrogen concentration sensor
positioned proximate an intake valve. The plurality of sensors (102) provides
additional combustion and timing feedback. The crank-angle position sensor10
enables cycle-resolved data alignment. The ionization or knock sensor provides
combustion heat and timing indicators. The hydrogen concentration sensor provides
localized fuel presence information. The combined sensor set enhances
observability of pre-ignition conditions.
[0036] The predictive backfire prevention system (100) further comprises a15
processing unit (104) communicatively coupled to the plurality of sensors (102).
The processing unit (104) is configured to receive and process the real-time
combustion-related data. In some embodiments, the processing unit (104) operates
on a cycle-by-cycle basis synchronized with engine rotation. The processing unit
(104) computes a set of normalized feature values derived from the sensor data. The20
set of normalized feature values includes an estimate of residual gas temperature.
The feature set further includes a local hydrogen equivalence ratio and a manifold
pressure rate-of-change. Normalization enables consistent feature scaling across
operating conditions.
[0037] In some embodiments, the estimate of residual gas temperature is computed25
as a function of measured intake manifold temperature and at least one combustion
heat indicator. The combustion heat indicator may include ionization amplitude,
recent ignition timing information, or deviation of intake manifold pressure from
atmospheric pressure. The residual gas temperature estimate reflects thermal energy
retained from prior combustion cycles. This estimation enables assessment of hot-30
spot induced ignition risk. The processing unit (104) updates the estimate
11
dynamically. The estimation is performed without requiring direct in-cylinder
temperature sensing. The approach supports practical implementation.
[0038] The processing unit (104) is further configured to compute a raw backfire
risk score using a weighted combination of the normalized feature values. In some
embodiments, the weighted combination is implemented using a linear model. The5
linear model assigns weighting coefficients to each normalized feature value. The
raw backfire risk score represents a combined influence of thermal, fueling, and
dynamic engine conditions. The processing unit (104) maps the raw backfire risk
score through a logistic sigmoid function. The mapping generates a predictive
backfire risk index representing a probability of intake backfire. The predictive10
backfire risk index is continuously updated.
[0039] In some embodiments, the processing unit (104) applies a sliding window
across a plurality of consecutive combustion cycles. The predictive backfire risk
index values within the sliding window are evaluated collectively. The processing
unit (104) compares the predictive backfire risk index to one or more predefined15
thresholds. Corrective action is triggered only when the predictive backfire risk
index exceeds a threshold for a predefined persistence duration. This persistence
logic reduces false positives. The sliding window enables temporal filtering of
transient conditions. The approach ensures stability of control decisions. The
predictive backfire prevention system (100) balances responsiveness and20
robustness.
[0040] The predictive backfire prevention system (100) further comprises a control
interface (106) configured to communicate corrective control requests. When the
predictive backfire risk index exceeds at least one predefined threshold, the control
interface (106) commands at least one corrective engine control action. In some25
embodiments, the corrective engine control action comprises temporarily retarding
ignition timing by a variable crank-angle amount. The magnitude of retard may be
proportional to the predictive backfire risk index. The ignition timing adjustment
reduces residual gas temperature in subsequent cycles. The action is applied
proactively before backfire occurs. The control interface (106) supports graded30
response.
12
[0041] In some embodiments, the corrective engine control action comprises
reducing or suspending hydrogen fuel injection for a predetermined duration. The
reduction or suspension purges hydrogen from an intake manifold volume. This
action moves local conditions outside a flammability range. The predetermined
duration may vary based on risk severity. The corrective action may be combined5
with ignition timing adjustment. The control strategy prevents accumulation of
unburned hydrogen near hot surfaces. The corrective action is temporary and
reversible. Normal operation resumes after recovery conditions are met.
[0042] The processing unit (104) may further comprise an adaptive learning module
(108) configured to refine backfire risk estimation. In some embodiments, the10
adaptive learning module (108) updates weighting coefficients of the raw backfire
risk score. The update is based on tagged operating data preceding a detected
backfire or near-backfire event. The tagged data represents high-risk operating
patterns. The adaptive learning module (108) enables compensation for engine
aging. The module further compensates for hydrogen fuel composition variation,15
ambient environmental conditions, or sensor drift. Learning occurs progressively
during field operation. Manual recalibration is reduced.
[0043] In some embodiments, the processing unit (104) communicates with a main
engine control unit via a vehicle communication bus. The processing unit (104)
transmits backfire risk status and corrective action requests. The main engine20
control unit retains final authority over actuator execution. This architecture allows
integration without disrupting existing control hierarchies. The processing unit
(104) functions as an intelligent advisory and preventive layer. The engine control
unit validates and executes commands within safety limits. The communication
enables coordination with fuel, ignition, and airflow control systems. The25
architecture supports modular deployment. The predictive backfire prevention
system (100) enhances safety while preserving engine performance.
[0044] FIG. 2 illustrates a flowchart showing a method (200) for operating the
predictive backfire prevention system (100), according to an embodiment of the
present disclosure.30
13
[0045] At operation 202, the plurality of sensors (102) is configured to generate
real-time combustion-related data including at least intake manifold pressure, intake
manifold temperature, engine speed, and hydrogen fuel delivery information. The
plurality of sensors (102) operate continuously across varying engine loads and
speeds. The sensor data represents instantaneous intake and combustion conditions.5
The plurality of sensors (102) collectively enable real-time observation of
parameters relevant to backfire occurrence. The predictive backfire prevention
system (100) utilizes existing engine sensing infrastructure.
[0046] The plurality of sensors (102) may further include one or more additional
sensing elements disposed at selected engine locations. In some embodiments, the10
plurality of sensors (102) comprises at least one of an ionization sensing circuit, a
knock sensor, a crank-angle position sensor, or a hydrogen concentration sensor
positioned proximate an intake valve. The plurality of sensors (102) provides
additional combustion and timing feedback. The crank-angle position sensor
enables cycle-resolved data alignment. The ionization or knock sensor provides15
combustion heat and timing indicators. The hydrogen concentration sensor provides
localized fuel presence information. The combined sensor set enhances
observability of pre-ignition conditions.
[0047] At operation 204, the processing unit (104) is communicatively coupled to
the plurality of sensors (102) and is configured to compute, for each combustion20
cycle, a set of normalized feature values derived from the real-time combustion-
related data. The set of normalized feature values includes at least an estimate of
residual gas temperature, a local hydrogen equivalence ratio, and a manifold
pressure rate-of-change. The feature set further includes a local hydrogen
equivalence ratio and a manifold pressure rate-of-change. Normalization enables25
consistent feature scaling across operating conditions.
[0048] At operation 206, the processing unit (104) is configured to compute a raw
backfire risk score using a weighted combination of the set of normalized feature
values. In some embodiments, the weighted combination is implemented using a
linear model. The linear model assigns weighting coefficients to each normalized30
14
feature value. The raw backfire risk score represents a combined influence of
thermal, fueling, and dynamic engine conditions.
[0049] At operation 208, the processing unit (104) is configured to map the raw
backfire risk score to a predictive backfire risk index representing a probability of
an intake backfire event. The processing unit (104) maps the raw backfire risk score5
through a logistic sigmoid function. The predictive backfire risk index is
continuously updated.
[0050] At operation 210, the processing unit (104) is configured to compare the
predictive backfire risk index to one or more predefined thresholds.
[0051] At operation 212, the control interface (106) is configured to, when the10
predictive backfire risk index exceeds at least one of the predefined thresholds,
command at least one corrective engine control action selected from hydrogen fuel
injection modification, ignition timing adjustment, or temporary hydrogen fuel cut-
off. Further, the predictive backfire risk index is computed prior to occurrence of a
backfire event and the corrective engine control action is applied proactively to15
suppress intake backfire.
[0052] Various embodiments of the present invention provide notable advantages
through a predictive backfire prevention system (100) comprising an engine control
unit, a plurality of engine sensors, a backfire risk computation module, and a
coordinated control module configured to operate as an integrated software-based20
safety architecture. The engine control unit continuously executes a dynamic
weighted backfire risk index computation using fused inputs from the plurality of
engine sensors, enabling real-time and continuous risk estimation. The backfire risk
computation module supports predictive and preemptive control actions, thereby
preventing backfire events before physical ignition propagation occurs. A self-25
learning adaptive model embedded within the engine control unit updates
correlation weights based on operating history, improving accuracy over time
without manual recalibration. The coordinated control module proportionally
regulates injector timing, ignition timing, and airflow based on computed risk levels.
Predictive logging and trend analysis are performed within the engine control unit30
to support diagnostic intelligence and preventive maintenance. The predictive
15
backfire prevention system (100) is purely software-implemented using existing
hardware infrastructure, reducing cost and complexity. Overall, the integrated
architecture enhances combustion safety, operational stability, and performance
reliability of hydrogen internal combustion engines across diverse operating
conditions.5
[0053] It has thus been seen the predictive backfire prevention system (100) for the
hydrogen-fueled internal combustion engine, as described. The predictive backfire
prevention system (100) in any case could undergo numerous modifications and
variants, all of which are covered by the same innovative concept; moreover, all of
the details can be replaced by technically equivalent elements. In practice, the10
components used, as well as the numbers, shapes, and sizes of the components can
be whatever according to the technical requirements. The scope of protection of the
invention is therefore defined by the attached claims.

CLAIMS

1. A predictive backfire prevention system (100) for a hydrogen-fueled internal
combustion engine, comprising:
a plurality of sensors (102) configured to generate real-time
combustion-related data including at least intake manifold pressure, intake
manifold temperature, engine speed, and hydrogen fuel delivery information;
a processing unit (104) communicatively coupled to the plurality of
sensors (102) and configured to:
compute, for each combustion cycle, a set of normalized
feature values derived from the real-time combustion-related data, the
set of normalized feature values including at least an estimate of
residual gas temperature, a local hydrogen equivalence ratio, and a
manifold pressure rate-of-change;
compute a raw backfire risk score using a weighted
combination of the set of normalized feature values;
map the raw backfire risk score to a predictive backfire risk
index representing a probability of an intake backfire event; and
compare the predictive backfire risk index to one or more
predefined thresholds; and
a control interface (106) configured to, when the predictive backfire
risk index exceeds at least one of the predefined thresholds, command at least
one corrective engine control action selected from hydrogen fuel injection
modification, ignition timing adjustment, or temporary hydrogen fuel cut-off,
wherein the predictive backfire risk index is computed prior to
occurrence of a backfire event and the corrective engine control action is
applied proactively to suppress intake backfire.
2. The predictive backfire prevention system (100) as claimed in claim 1,
wherein the plurality of sensors (102) further comprises at least one of an
ionization sensing circuit, a knock sensor, a crank-angle position sensor, or a
hydrogen concentration sensor disposed proximate an intake valve.
3. The predictive backfire prevention system (100) as claimed in claim 1,
wherein the estimate of residual gas temperature is computed as a function of
measured intake manifold temperature and at least one combustion heat
indicator selected from ionization amplitude, recent ignition timing, or
manifold pressure deviation from atmospheric pressure.
4. The predictive backfire prevention system (100) as claimed in claim 1,
wherein the raw backfire risk score is computed using a linear model and
wherein the predictive backfire risk index is generated by applying a logistic
sigmoid function to the raw backfire risk score.
5. The predictive backfire prevention system (100) as claimed in claim 1,
wherein the processing unit (104) applies a sliding window across a plurality
of consecutive combustion cycles and triggers the corrective engine control
action only when the predictive backfire risk index exceeds a threshold for a
predefined persistence duration.
6. The predictive backfire prevention system (100) as claimed in claim 1,
wherein the corrective engine control action comprises temporarily retarding
ignition timing by a variable crank-angle amount proportional to the
predictive backfire risk index.
7. The predictive backfire prevention system (100) as claimed in claim 1,
wherein the corrective engine control action comprises reducing or
suspending hydrogen injection for a predetermined duration to purge
hydrogen from an intake manifold volume.
8. The predictive backfire prevention system (100) as claimed in claim 1,
wherein the processing unit (104) further comprises an adaptive learning
module (108) configured to update weighting coefficients of the raw backfire
risk score based on tagged operating data preceding a detected backfire or
near-backfire event and wherein the adaptive learning module (108) is
configured to refine the predictive backfire risk index to compensate for at
least one of engine aging, hydrogen fuel composition variation, ambient
environmental conditions, or sensor drift.
9. The predictive backfire prevention system (100) as claimed in claim 1,
wherein the processing unit (104) communicates with a main engine control
unit via a vehicle communication bus and transmits backfire risk status and
corrective action requests while the main engine control unit retains final
authority over actuator execution.
10. A method (200) for operating a predictive backfire prevention system (100)
for a hydrogen-fueled internal combustion engine, the method (200)
comprising:
generating, via a plurality of sensors (102), real-time combustion-
related data including at least intake manifold pressure, intake manifold
temperature, engine speed, and hydrogen fuel delivery information;
computing, via a processing unit (104) communicatively coupled to
the plurality of sensors (102), for each combustion cycle, a set of normalized
feature values derived from the real-time combustion-related data, the set of
normalized feature values including at least an estimate of residual gas
temperature, a local hydrogen equivalence ratio, and a manifold pressure rate-
of-change;
computing, via the processing unit (104), a raw backfire risk score
using a weighted combination of the set of normalized feature values;
mapping, via the processing unit (104), the raw backfire risk score to
a predictive backfire risk index representing a probability of an intake backfire
event; and
comparing, via the processing unit (104), the predictive backfire risk
index to one or more predefined thresholds; and
commanding, via a control interface (106), when the predictive
backfire risk index exceeds at least one of the predefined thresholds, at least
one corrective engine control action selected from hydrogen fuel injection
modification, ignition timing adjustment, or temporary hydrogen fuel cut-off,
wherein the predictive backfire risk index is computed prior to
occurrence of a backfire event and the corrective engine control action is
applied proactively to suppress intake backfire.