Description:Complete Specification:
The following specification describes and ascertains the nature of this invention and the manner in which it is to be performed

Field of the invention
[0001] The present disclosure relates to the field of automotive steering. In particular the invention discloses a method to calculate a steering angle using a trained AI model during a clutch slip condition and a Electronic Control Unit (ECU) configured to do the same.

Background of the invention
[0002] The steering angle measurement determines where the driver wants to steer, and vehicle systems match the steering wheel with the vehicle's wheels. Modern day vehicles equipped with features such as the Highly automated Driving (HAD), Electronic Stability Program (ESP), Adaptive cruise control (ACC), Park assist etc. require steering-angle information. Conventional steering angle sensors are mounted on steering column or the steering rack, the steering column is mechanically coupled to a steering rack by means of a rack and pinion arrangement.

[0003] The steering rack further comprises a motor and rack position sensor (RPS). The RPS detects a linear position of a steering rack. Modern vehicles use power steering or an Electronic Power Steering (EPS) system that reduce a driver's effort to turn steering wheel of a vehicle. Commonly used EPS systems control and assist the steering system with the support of this intelligent electric motor in the steering rack. Based on the signal from the steering angle sensor, a control unit calculates an optimal steering support and sends the information to the electric motor to provide the necessary assistance.

[0004] This electric motor is coupled to the rack by means of a worm wheel assembly and a secondary pinion. The worm wheel assembly comprises a clutch that is attached to the pinion. However, this clutch detaches from the pinion when disproportionate force acts upon the steering rack. This condition is known as the “clutch slip”. In conventional EPS systems during the condition of clutch slip the synchronization between the steering angle sensor and rotor position sensor is lost. Hence measurement of steering angle is deemed invalid and hence features such as the Highly automated driving (HAD), Electronic Stability Program (ESP), Park assist etc. that require steering-angle information are disabled. The present invention aims to calculate the value of steering angle even during the clutch slip condition using a trained AI model.

Brief description of the accompanying drawings
[0005] An embodiment of the invention is described with reference to the following accompanying drawings:
[0006] Figure 1 depicts an automotive Electronic Power steering (EPS) system (100);
[0007] Figure 2 depicts a portion of steering rack (106) with a worm wheel assembly (120);
[0008] Figure 3 illustrates method steps to train an AI model (140) to calculate a magnitude of clutch slip;
[0009] Figure 4 illustrates method steps to calculate steering angle for the EPS system (100).

Detailed description of the drawings
[0010] Figure 1 depicts an automotive Electronic Power steering (EPS) system. The EPS system (100) comprises a steering wheel (102) coupled to a steering rack (106) through a steering column (104). The steering column (104) comprises an input shaft from the steering connected to an output shaft by means of a torsion bar and a gear system. The input shaft translates the rotational motion on the steering wheel (102) to a telescopic movement of the torsion bar and a gear system. A steering angles sensor is mechanically coupled to the steering column (104). The output shaft of the steering column (104) is mechanically coupled to a steering rack (106) (steering rack (106)) by means of a steering rack (106) and pinion arrangement.

[0011] Figure 2 depicts a portion of the steering rack (106) with a worm wheel assembly (120). A motor (108) is coupled to the steering rack (106) by means of a worm wheel assembly (120) and a secondary pinion (121). The worm wheel assembly (120) comprises a worm (122), worm wheel (123) and at least a clutch (126) that is attached to the pinion under normal circumstances. This clutch (126) detaches from the pinion when disproportionate force acts upon the steering rack (106). This condition is known as clutch slip. Sensors such as the rotor position sensor (RPS) are mounted on the steering rack (106). An Electronic control unit (ECU (130)) in communication with the steering angle sensor (110) and other components of the EPS system (100).

[0012] The ECU (130) is logic circuitry and software programs that respond to and processes logical instructions to get a meaningful result. Modern day vehicles may contain a plurality of control unit s like the Airbag control unit, Transmission control unit, Glow time control unit, Heating control unit, Vehicle charge communication unit, Engine control unit, Vehicle control unit, Steering control unit and the like. Each control unit coordinates the components specific to them for example an Engine control unit can provide torque coordination, operation, and gearshift strategies, on board diagnosis, monitoring, thermal management and much more for electrified and connected powertrains. The ECU (130) used here may be a Steering control unit.

[0013] The ECU (130) may be implemented in the system as one or more microprocessors, microcomputers, microcontrollers, digital signal processors, central processing units, state machines, logic circuitries, one or more microchips or integrated circuits interconnected using a parent board, hardwired logic, software stored by a memory device and executed by a microprocessor, firmware, an application specific integrated circuit (ASIC), and/or a field programmable gate array (FPGA), and/or any component that operates on signals based on operational instructions.

[0014] The ECU (130) is in communication with a trained AI model (140). An AI model with reference to this disclosure can be explained as a component which runs a model. A model can be defined as reference or an inference set of data, which uses different forms of correlation matrices. Using these models and the data from these models, correlations can be established between different types of data to arrive at some logical understanding of the data. A person skilled in the art would be aware of the different types of AI models such as linear regression, naïve bayes classifier, support vector machine, neural networks and the like. A person skilled in the art will also appreciate that the AI model (140) may be implemented as a set of software instructions or a hardware (such as neural network chips) or a combination of software and hardware of the same.

[0015] Some of the typical tasks performed by AI models are classification, clustering, regression etc. Majority of classification tasks depend upon labeled datasets; that is, the data sets are labelled manually in order for a neural network to learn the correlation between labels and data. This is known as supervised learning. Some of the typical applications of classifications are: face recognition, object identification, gesture recognition, voice recognition etc. Clustering or grouping is the detection of similarities in the inputs. The cluster learning techniques do not require labels to detect similarities. Learning without labels is called unsupervised learning. The AI model (140) used here in trained in a supervised manner in accordance with method steps 200. In an exemplary embodiment of the present invention the AI model (140) used is a neural network.

[0016] The ECU (130) configured is perform method steps in accordance with figure 3. It is configured to determine a value of a worm wheel torque and rotor acceleration; receive a value of a steering angle before the clutch slip from a sensor on the steering column (104); feed the value of worm wheel torque and rotor acceleration as input to a trained AI model (140); execute the AI model (140) to determine a magnitude of clutch slip; calculate an instantaneous value of steering angle by adjusting the received value of steering angle in accordance with the magnitude of clutch slip. The instantaneous value of steering angle is calculated for the clutch slip condition.

[0017] It should be understood at the outset that, although exemplary embodiments are illustrated in the figures and described below, the present disclosure should in no way be limited to the exemplary implementations and techniques illustrated in the drawings and described below.

[0018] Figure 3 illustrates method steps to train an AI model (140) to calculate a magnitude of clutch slip. The AI model (140) when trained is part of the EPS system (100) as depicted in accordance with figure 1. The AI model (140) training happens in a test set-up. The set-up comprises the EPS system (100) i.e. a steering wheel (102) coupled to a steering rack (106) through a steering column (104), a motor (108) coupled to the steering rack (106) by means of a worm wheel assembly (120) and a secondary pinion (121), the worm wheel assembly (120) comprising a clutch (126) that detaches from the pinion when disproportionate force acts upon the steering (106). Additionally, the EPS system (100) in the test set-up is connected to sensor that provides the real-time steering angle even during the clutch slip phase (when the clutch (126) detaches from the secondary pinion (121)).

[0019] Method step 201 comprises injecting disproportionate force on the steering rack (106) such that a clutch slip condition occurs in the EPS system (100).

[0020] Method step 202 comprises determining a value of a worm wheel torque and rotor acceleration post injection of the disproportionate force by means of an Electronic control unit (ECU (130)). The ECU (130) is part of the EPS system (100) and in communication with various sensors of the EPS system (100). The ECU (130) is further in communication with AI model (140).

[0021] Method step 203 comprises feeding the value of worm wheel torque and rotor acceleration as input to the AI model (140). Method step 203 comprises determining a loss function of the input parameters for the magnitude of clutch slip. Determining the loss function further comprises, first receiving a value of steering angle just before the clutch slip from a sensor on the steering column (104). This is followed by recording an actual value of steering angle during the clutch slip by means of the additional sensor connected to the EPS. The additional sensor in essence gives a ground truth value of the steering angle when the clutch (126) is slipping and the steering angle sensor (110) is inactive. This additional sensor will not be available in the series vehicle, it is exclusive to the Test setup used by us for supervised training of the AI model (140). The loss function is determined as the difference between the received value (i.e. value of steering angle just before the clutch slip measured by the sensor mounted on the steering column (104)) and recorded value ( ground truth value).

[0022] Method steps 204 comprises optimizing the network parameters and hyperparameters of the AI model (140) to minimize the loss function. Neural networks are inspired by the biological neural network or brain cell i.e. neurons. The network parameters include but are not limited to a layers, filter and the like. For simplicity, in computer science, a network of neurons are represented as a set of layers. These layers are categorized into three classes which are input, hidden, and output. Every network has a single input layer and a single output layer. Different layers perform different kinds of transformations/operations on their inputs. Data flows through the network starting at the input layer and moving through the hidden layers until the output layer is reached. Hyperparameter is a parameter whose value is used to control the learning process. While networks parameters are learned during the training stage, hyper parameters are given/chosen. Hyper parameters are typically characterized by the learning rate, learning pattern and the batch size. In method step 204 we basically tune the hyperparameters and network parameters such that the loss function is minimal which means that the neural network learns to correlate the input parameters to give the accurate steering angle during clutch slip.

[0023] Figure 4 illustrates the method steps to calculate steering angle for the EPS system (100). The EPS system (100) and its components have elucidated in accordance with figure 1 and figure 2. For the purposes of clarity, it is reiterated that the EPS system (100) comprises a steering wheel (102) coupled to a steering rack (106) through a steering column (104), a motor (108) coupled to the steering rack (106) by means of a worm wheel assembly (120) and a secondary pinion (121), the worm wheel assembly (120) comprising a clutch (126) that detaches from the pinion when disproportionate force acts upon the steering rack (106). A steering angles sensor is mechanically coupled to the steering column (104). Sensors such as the rotor position sensor (RPS) , steering angle sensor are mounted on the steering rack (106). An Electronic control unit (ECU (130)) in communication with the steering angle sensor (110) and other components of the EPS system (100). The ECU (130) is in communication with an AI model (140) trained in accordance with method step 200. This trained model is now deployed real-time in the EPS system (100). In an exemplary embodiment of the present invention, the AI model (140) is a neural network.

[0024] Method step 301 comprises determining a value of a worm wheel torque and rotor acceleration by means of the Electronic control unit (ECU (130)). Worm wheel torque and rotor acceleration are parameters determined by the ECU (130) in communication with the rotor position sensor and the motor (108).

[0025] Method step 302 comprises receiving a value of a steering angle before the clutch slip from the sensor on the steering column (104). The last valid value of the steering angle (i.e. before the clutch slip condition) recorded by the steering angle sensor (110) is taken into consideration.

[0026] Method step 303 comprises feeding the value of worm wheel torque and rotor acceleration as input to trained AI model (140). Method step 304 comprises executing the AI model (140) to determine a magnitude of clutch slip. We have seen in accordance with method steps 200 the AI model (140) has learnt the correlation between magnitude of slip of the input parameters.

[0027] Method step 305 comprises calculating an instantaneous value of steering angle by adjusting the received value of steering angle in accordance with the magnitude of clutch slip. The instantaneous value of steering angle is calculated for the clutch slip condition.

[0028] A person skilled in the art will appreciate that while these method steps describes only a series of steps to accomplish the objectives, these methodologies may be implemented with modification and customizations to the EPS system (100). This idea to develop a method to calculate steering angle during a clutch slip condition using a trained AI model (140) ensures availability of the system functions such as Highly automated Driving (HAD), Electronic Stability Program (ESP), Park assist etc. even in clutch slip condition.

[0029] It must be understood that the embodiments explained in the above detailed description are only illustrative and do not limit the scope of this invention. Any modification to the method to calculate a steering angle using the trained AI model (140) during a clutch slip condition and an Electronic Control Unit (ECU (130)) configured to do the same are envisaged and form a part of this invention. The scope of this invention is limited only by the claims.
, Claims:We Claim:
1. A method (300) to calculate steering angle for an Electronic Power Steering (EPS) system, said EPS system (100) comprising a steering wheel (102) coupled to a steering rack (106) through a steering column (104), a motor (108) coupled to the steering rack (106) by means of a worm wheel assembly (120) and a secondary pinion (121), the worm wheel assembly (120) comprising a clutch (126) that detaches from the pinion (clutch slip) when disproportionate force acts upon the steering rack (106), the method comprising:
Determining (301) a value of a worm wheel torque and rotor acceleration by means of a steering control unit (SCU);
receiving (302) a value of a steering angle before the clutch slip from a sensor on the steering column (104);
feeding (303) the value of worm wheel torque and rotor acceleration as input to trained AI model (140);
executing (304) the AI model (140) to determine a magnitude of clutch slip;
calculating (305) an instantaneous value of steering angle by adjusting the received value of steering angle in accordance with the magnitude of clutch slip.

2. The method (300) to calculate steering angle as claimed in claim 1, wherein the trained AI model (140) is a neural network.

3. The method (300) to calculate steering angle as claimed in claim 1, wherein the instantaneous value of steering angle is calculated for the clutch slip condition.

4. A method (200) to train an AI model (140) to calculate a magnitude of clutch slip in an Electronic Power Steering (EPS) system, said EPS system (100) comprising a steering wheel (102) coupled to a steering rack (106) through a steering column (104), a motor (108) coupled to the steering rack (106) by means of a worm wheel assembly (120) and a secondary pinion (121), the worm wheel assembly (120) comprising a clutch (126) that detaches from the pinion when disproportionate force acts upon the steering rack (106), the method comprising:
injecting (201) said disproportionate force on the steering rack (106) for a clutch slip condition;
determining (202) a value of a worm wheel torque and rotor acceleration post injection of the disproportionate force by means of a steering control unit (SCU);
deeding (203) the value of worm wheel torque and rotor acceleration as input to the AI model (140);
determining (204) a loss function of the input parameters for the magnitude of clutch slip;
optimizing (205) the network parameters and hyperparameters of the AI model (140) to minimize the loss function.

5. The method (200) to train an AI model (140) to calculate a magnitude of clutch slip as claimed in claim 4, wherein the determining the loss function further comprises:
receiving a value of steering angle just before the clutch slip from a sensor on the steering column (104);
recording an actual value of steering angle during the clutch slip by means of an additional sensor connected to the EPS;
determining the loss function as difference between the received value and recorded value.

6. An Electronic Control Unit (ECU (130)) for an Electronic Power Steering (EPS) system, said EPS system (100) comprising a steering wheel (102) coupled to a steering rack (106) through a steering column (104), a motor (108) coupled to the steering rack (106) by means of a worm wheel assembly (120) and a secondary pinion (121), the worm wheel assembly (120) comprising a clutch (126) that detaches from the pinion when disproportionate force acts upon the steering rack (106), the ECU (130) configured to:
determine a value of a worm wheel torque and rotor acceleration;
receive a value of a steering angle before the clutch slip from a sensor on the steering column (104);
feed the value of worm wheel torque and rotor acceleration as input to a trained AI model (140);
execute the AI model (140) to determine a magnitude of clutch slip;
calculate an instantaneous value of steering angle by adjusting the received value of steering angle in accordance with the magnitude of clutch slip.

7. The Electronic Control Unit (ECU (130)) for an EPS system (100) as claimed in claim 6, wherein the instantaneous value of steering angle is calculated for the clutch slip condition.

8. The Electronic Control Unit (ECU (130)) for an EPS system (100) as claimed in claim 6, wherein the trained AI model (140) is a neural network.