Systems and methods for detonation control in spark ignition engines using analog neuromorphic computing hardware

1. An apparatus for detonation control in spark ignition engines, the apparatus comprising:

an analog neuromorphic hardware device trained to detect early detonation in cylinders of a spark ignition engine;

a knock sensor coupled to the spark ignition engine;

an ignition coil for the spark ignition engine; and

an Electronic Control Unit (ECU) for the spark ignition engine;

wherein the analog neuromorphic hardware device is configured to:

receive knock sensor signals from the knock sensor;

receive ignition coil data from the ignition coil;

determine a knock level and ignition quality measure based on the received knock sensor signals and the received ignition coil data; and

transmit the knock level and ignition quality measure to the ECU.

2. The apparatus of claim 1 , wherein transmitting the knock level and ignition quality measure to the ECU is performed periodically at a rate matching crankshaft rotation of the spark ignition engine.

3. The apparatus of claim 1 , wherein the ECU controls ignition timing for the spark ignition engine according to the knock level and ignition quality measure transmitted from the analog neuromorphic hardware device.

4. The apparatus of claim 1 , wherein the ECU controls closer-to-knock regimes of ignition timing for the spark ignition engine according to the knock level and ignition quality measure transmitted from the analog neuromorphic hardware device.

5. The apparatus of claim 1 , wherein the ECU controls fuel mixtures in cylinders of the spark ignition engine according to the knock level and ignition quality measure transmitted from the analog neuromorphic hardware device.

6. The apparatus of claim 1 , wherein the analog neuromorphic hardware device is trained to detect the knock level based on a shape of the knock sensor signals.

7. The apparatus of claim 1 , wherein the knock sensor signals include noise due to transmission, bodywork, and/or age of the spark ignition engine, and the analog neuromorphic hardware device is trained using data that includes such noise.

8. The apparatus of claim 1 , wherein knock sensor signals are irregular, and the analog neuromorphic hardware device determines the knock level and ignition quality measure dynamically based on the received knock sensor signals and the received ignition coil data during a predefined period of time.

9. The apparatus of claim 1 , wherein the analog neuromorphic hardware device is integrated within the ECU.

10. The apparatus of claim 1 , further comprising an analog-to-digital converter for converting the knock level and ignition quality measure from the analog neuromorphic hardware device to a digital form before transmitting to the ECU.

11. The apparatus of claim 1 , wherein the analog neuromorphic hardware device implements a trained Convolutional Neural Network that is trained to (i) accept the knock sensor signals and the ignition coil data in real time and (ii) output detonation level on a log scale.

12. The apparatus of claim 1 , wherein the analog neuromorphic hardware device implements a trained Convolutional Neural Network that extracts knock sensor sound patterns that are correlated with detonation level and pre-detonation states.

13. The apparatus of claim 1 , wherein the analog neuromorphic hardware device implements a trained neural network that is trained using datasets of knock sensor and ignition coil data, obtained during real engine operation, including pre-detonation, beginning of detonation, and full detonation.

14. The apparatus of claim 1 , wherein the analog neuromorphic hardware device implements a trained neural network that is trained to process knock sensor and ignition coil signals and classify them as detonation level and ignition level.

15. The apparatus of claim 1 , wherein the analog neuromorphic hardware device is fabricated by a series of steps comprising:

obtaining a neural network topology and weights of a trained neural network model;

transforming the neural network topology into an equivalent analog network of analog components;

computing a weight matrix for the equivalent analog network based on the weights of the trained neural network model, wherein each element of the weight matrix represents one or more connections between analog components of the equivalent analog network;

generating a schematic model for implementing the equivalent analog network based on the weight matrix, including selecting component values for the analog components; and

fabricating an integrated circuit, according to the schematic model, using a lithographic process.

16. The apparatus of claim 15 , wherein generating the schematic model includes generating a resistance matrix for the weight matrix, wherein each element of the resistance matrix corresponds to a respective weight of the weight matrix and represents a resistance value.

17. The apparatus of claim 1 , wherein the analog neuromorphic hardware device is an integrated circuit comprising an analog network of analog components configured to implement a trained neural network model that is trained to predict knock level and ignition quality measure based on the knock sensor signals and ignition coil data.

18. A method for detonation control in spark ignition engines, the method comprising:

receiving knock sensor signals from a knock sensor coupled to a spark ignition engine, the knock sensor signals including noise due to transmission, bodywork, and/or spark ignition engine age;

receiving ignition coil data from an ignition coil for the spark ignition engine;

using an analog neuromorphic hardware device trained using data that includes noise due to transmission, bodywork, and/or spark ignition engine age, determining knock level and ignition quality measure based on the received knock sensor signals and the received ignition coil data; and

transmitting the knock level and ignition quality measure to an Electronic Control Unit (ECU) for the spark ignition engine.

19. The method of claim 18 , wherein transmitting the knock level and ignition quality measure to the ECU is performed periodically at a rate matching crankshaft rotation of the spark ignition engine.

20. An apparatus for detonation control in spark ignition engines, the apparatus comprising:

a knock sensor coupled to a spark ignition engine;

an ignition coil for the spark ignition engine;

an Electronic Control Unit (ECU) for the spark ignition engine; and

an analog neuromorphic hardware device implementing a trained Convolutional Neural Network that extracts knock sensor sound patterns that are correlated with detonation level and pre-detonation states;

wherein the analog neuromorphic hardware device is configured to:

receive knock sensor signals from the knock sensor;

receive ignition coil data from the ignition coil;

determine a knock level and ignition quality measure based on the received knock sensor signals and the received ignition coil data; and

transmit the knock level and ignition quality measure to the ECU.