WO2008139493A1

WO2008139493A1 - Configurable semiconductor virtual sensor

Info

Publication number: WO2008139493A1
Application number: PCT/IN2008/000294
Authority: WO
Inventors: Dinesh Shyamsundar
Original assignee: Ucal Fuel Systems Limited
Priority date: 2007-05-11
Filing date: 2008-05-12
Publication date: 2008-11-20

Abstract

The subject matter described herein is directed towards a virtual sensor device having one or more processing modules, at least a memory and at least one port. The port communicates between the virtual sensor device and a processor of an electronic control unit. The memory stores configurable coded instructions for estimating variable properties associated with an object or process. The configurable coded instructions render the virtual sensor device compatible with the processor. The configurable coded instructions make the virtual sensor platform independent.

Description

TECHNICAL FIELD

The subject matter described herein in general relates to virtual sensors and in particular relates to platform independent configurable semiconductor virtual sensor.

BACKGROUND

In a control environment, various actuators function with the help of physical sensors and input signals. The sensors and inputs are controlled typically by an Electronic Controller Unit (ECU). With the advancement in technology, various sensors have been replaced by a set of instruction codes residing in the ECU. These instruction codes perform the function of manipulating the ECU to generate signals similar to the generation of a sensor signal. These instruction codes are referred to as virtual sensors.

Virtual sensors do not have a sensing element. Virtual sensors estimate variable properties or states associated with an object or process using statistical models. Virtual sensors are employed either in conjunction with physical sensors or independently. Virtual sensors may also utilize the readings of other physical sensors, estimating variable properties or state of an object or process, to compute the variable properties or state of other objects or processes.

Virtual sensors can replace physical sensors, performing the same function and eliminating various shortcoming associated with the physical sensor. A physical sensor, in a fast transient environment is incapable of measuring all possible data points due to its inertia in response. In many cases, it is desirable to have a sensing mechanism that accepts to the fast transient mode and is capable of instantaneous output.

Additionally physical sensor may be too far downstream, physically, to provide timely information for process adjustments. This time lag can introduce errors in the measurement which hamper efficient control and operator response. Further, in many cases if the sensor environment is hostile, the parameter of serviceability and accessibility may be and cost may be very high, thereby inhibiting its widespread use. Virtual sensors are built as models which are coded and then the object files of the coded models are linked into the ECU. While virtual sensors reduce the input/output (I/O) count of the microcontroller and enhance reliability, there are some associated problems. The process of building a model, coding it and linking the object files of the model to the ECU software requires large program memory space. Furthermore, the implementation of a virtual sensor is dependent on the microcontroller present in the ECU. Conventional virtual sensor models have to be recompiled for eveiy new target system making the sensor dependant on the programming of the microcontroller leaving little for the end users and OEM for modifications.

Moreover, conventional virtual sensors occupy a large memory space. Increase in the number of virtual sensors associated with a system, increases the complexity of the system. Even replication of the functionality for similar end use, the entire algorithm requires the recoding of the models for different processors. This process is cumbersome, and leaves room for errors.

SUMMARY

The configurable coded instructions eliminate the need for coding a new set of instructions and linking the object files associated with the new instructions to any new platform associated with the processor of the ECU. This saves a large amount of program memory space and lime. Furthermore, the implementation of a virtual sensor is independent of the processor present in the ECU.

These, and other features, aspects, and advantages of the present subject matter will become better understood with reference to the following description and appended claims. This summary is provided to introduce a selection of concepts in a simplified form. This summary is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used to limit the scope of the claimed subject matter.

BRIEF DESCRIPTION OF DRAWINGS

The above and other features, aspects, and advantages of the subject matter will become better understood with regard to the following description, appended claims, and accompanying drawings where:

Fig. I illustrates a block diagram illustrating an exemplary application architecture implementing a configurable virtual sensor device in the ECU.

Fig. 2 illustrates a block diagram illustrating components of a virtual sensor device.

Fig. 3 illustrates a block diagram illustrating components of a configurable virtual sensor device implemented on a semiconductor. .

DETAILED DESCRIPTION

Virtual sensors estimate variable properties or states associated with an object or process using statistical models employing physical sensors in conjunction with the virtual sensor or even without the physical sensors. The models can be implemented as an executable code within an ECU. Therefore virtual sensor code is stored in the memory of the ECU. To use these virtual sensors in different platforms would require the code for the virtual sensors to be recompiled and configured so as to conform to the platform requirements of the ECU. It would be gathered that a typical ECU can include at least one or more of a processor, memory, and a power supply module. In a typical ECU, the platform refers to one or more attributes of processors, such as, architecture, instruction sets, and so on.

To this end a configurable virtual sensor device is described that is platform independent i.e. independent of the coding of the processor it is interfaced with. The virtual sensor, is implemented as a device, such as on a semiconductor chip. The virtual sensor device includes a memory for storing instructions for estimating variable properties associated with one or more objects. Examples of such variable properties include temperature, pressure, and such.

The instructions present within the virtual sensor device are present in a precompiled format. The instruction in the virtual sensor device when interfaced with the processor of the ECU, would not require recoding or recompilation thus making, the virtual sensor device independent of the platform. The virtual sensor device interacts or interfaces with the processor through a data port. The data port allows data exchange between the virtual sensor device and the processor. The data port can be included but not limited to a serial or a parallel data port or any other similar ports. As the coding of the processor is independent of the instructions in the virtual sensor device, the virtual sensor device can be detached from the interfaced ECU and used for some other ECUs. In one implementation, the virtual sensor device can be configured either manually or remotely over a networked environment.

Fig. 1 illustrates a block diagram illustrating an exemplary architecture 100 implementing a configurable virtual sensor device. The architecture 100 includes a processor 102 associated with a memory 104 and one or more virtual sensor devices 106. The memory 104 includes any computer-readable medium known in the art, for example, volatile memory 104a and non-volatile memory 104b. The processor 102 can communicate with the virtual sensor devices 106 and memory 104 through ports 108, 110 and 112 for address and data. It would be noted that the manner in which the processor 102 exchanges data with the memory 104 through address and data bus 110, 112 is known in the art. The processor 102 can also be interfaced with signal conditioning circuits 114, output driver 116 and communications driver 118. These components assist in the general working of the architecture 100. The port 108 acts an interface allowing communication between the virtual sensor device 106 and the processor 102. In one implementation, the port 108 can be as serial port, a parallel port, or any such ports that are known in the art. The processor 102 may be implemented as one or more microprocessors, microcomputers, microcontrollers, dual core processors, and so forth. Among other capabilities, the processor 102 may be configured to fetch and execute computer readable instructions stored in a memory, such as memory 104.

The virtual sensor device 106 includes coded instructions that are utilized for estimating variable properties associated with one or more objects. The instructions are stored in the memory (not shown in Fig. 1) of the virtual sensor device 106. It would be appreciated that the memory of the virtual sensor device 106 is different from the memory 104. The coded instructions are precompiled. The precompiled code in the virtual sensor device 106 is independent of the instructions of the processor 102 and hence independent of the platform of the processor 102. In this manner, the instruction code of the virtual sensor device 106 need not conform to the coding of the processor 102. This is advantageous as any functionality can be implemented through coding instructions in the virtual sensor device 106 without the need for recompiling or recoding so as to make-the virtual sensor device 106 conform to the processor 102. The working of the virtual sensor device 106 in order to estimate one or more variable properties associated with the objects is further explained in conjunction with Fig. 2.

Fig. 2 indicates the components of the virtual sensor device 106. The virtual sensor device 106 includes an arithmetic logic unit (or ALU) 202 and a sensor memory depicted by non-volatile memory 204 and volatile memory 206. The non-volatile memory 204 (interchangeably referred to as sensor memory 204) stores instruction forming the logic of the virtual sensor. In one implementation, the sensor memory 204 includes control algorithm, virtual sensor algorithm, and configuration data, referred to as instructions 208. The instructions 208 stored in sensor memory 204 are precompiled. The instructions 208 circumvents the need to implement the code so as to conform to the code of a processor, such as processor 102, resulting in enabling an individual or consumer to use to virtual sensor device 106 with any type of ECU having a processor, such as processor 102,.

The ALU 202 and the instructions 208 can interact via the sensor address and the data buses 210, 212. The sensor address and the data buses 210, 212 allows the interaction of the components of the virtual sensor device 106 only, and are not interfaced to the address and data bus 110, 112 of the processor. The virtual sensor device 106 also includes a plurality of modules that are essential for the intended working of the virtual sensor device 106. Examples of such modules include, but are not limited to decoder 214, control unit 216, stack pointer 220, program counter 222. The data that is to be transmitted to the microcontroller of the ECU (e.g. microcontroller 102) can be transmitted through virtual sensor output 224, through communication driver 226. The virtual sensor output 224 interfaces with the port 108 for transmitting data between the virtual sensor device 106 and the processor 102.

The virtual sensor device 106 further includes a power source. In one implementation, the power source can be an external power source or internal to the virtual sensor device 106.

Fig. 3 illustrates a block diagram illustrating components of a configurable virtual sensor device implemented on a semiconductor. In one implementation, the virtual sensor device 106 can be implemented as a programmable system on chip (PSOC) 302. The PSOC 302 includes a interface module 304, communication module 306, and a macro FSM (Finite State Machine) module 308. The PSOC 302 communicates to the external world 303 through the interface module 304. In one implementation, the external world 303 is microprocessor, micro-controller based system with or without an operating system, and so on. The PSOC 302 can interface and communicate with any operating system and is protocol independent.

The macro FSM module 308 includes I/O buffers 310, RAM (Random Access Memory) module 312, and the micro FSM module 314. The micro FSM module 314 further includes a LUT (Look-Up-Table) module 316, mathematical module 318 and a cache module 320.

The macro FSM module 308 controls the interface module 304 and the communication module 306 apart from controlling the micro FSM module 314. The macro FSM module 308 interfaces with the external world 303 and accepts command requests and also controls the communication module 306 which in turn transfers input data from the external world 303 to the I/O buffers 310 of the macro FSM module 308. In the reverse mode, the macro FSM module 308 interfaces with the external world 303 and issues command requests and also controls communication module 306, which in turn transfers output data to the external world 303 from the I/O buffers 310 of the macro FSM module 308.

In one implementation, the RAM module 312 is a custom on-chip memory module and holds data for both the macro FSM module 308 and the micro FSM module 314. The micro FSM module 314 images the instructions in the virtual sensor device 106, such as instructions 208, to be executed. It is in principle the execution engine that uses the parameters in the LUT module 316, the mathematical module 318 and a cache module 320 to compute the output results. In one implementation, the LUT module 316 is a module interacting with a conventional look-up table, a linked list or a combination of both.

Although the subject matter has been described in detail with reference to certain preferred embodiments thereof, other embodiments are possible. As such, the spirit and scope of the appended claims should not be limited to the description of the preferred embodiment contained therein.

Claims

I/ We claim

1. Λ virtual sensor device comprising: one or more processing modules; at least a memory; at least one port, wherein said port communicate between said virtual sensor device and at least one processor; characterized in that said memory stores configurable coded instructions for estimating one or more parameters associated with an object or a process based on said coded instructions, rendering said virtual sensor device compatible with any processor.

2. The virtual sensor device of claim 1, wherein the processing module is an arithmetic logic unit.

3. The virtual sensor device of claim 1 , wherein said one or more parameters are from a group comprising variable properties or state, such as, temperature, pressure, vibration, oxygen content in exhaust and physical properties associated with said object or said process.

4. The virtual sensor device of claim I₃ wherein said device is implemented on a system-on-chip.

5. The virtual sensor device of claim 1, wherein the memory further includes configuration data for controlling execution of said configurable coded instructions.

6. Tlic virtual sensor device of claim 1 , powered through an external power source or internal to the virtual sensor device.

7. The semiconductor virtual sensor device of claim 1, wherein the port is a serial port.

8. The semiconductor virtual sensor device of claim 1, wherein the port is a parallel port.

9. The semiconductor virtual sensor device of claim 1 , wherein the port is a Wi-Fi port.

10. The semiconductor virtual sensor device of claim 1, wherein the configurable coded instructions are configured either manually or remotely.