TWM515677U - Parallel computation embedded sensing system - Google Patents

Description

Parallel computing embedded sensing system

This creation is about a sensing system, especially an embedded sensing system that is parallel computing.

With the advancement of technology, the existing monitoring system can process more diverse information. The monitoring system mainly includes a sensing module and an arithmetic unit. For example, the sensing module can sense temperature and illuminance. The sensing module of the environment change or the image capturing module, the computing unit executes a preset program to distinguish meaningful signals or meaningless noises from the sensing signals of the sensing module. In turn, it is possible to sense changes in the environment and adjust the actions of a controlled device. For example, the monitoring system can sense the ambient light intensity to determine whether to turn on the light fixture, or detect the ambient temperature to adjust the temperature of the helium gas, or perform image recognition by the computing unit to determine whether there is an intruder and determine the location of the intruder, or analyze one. Voice signals to determine the meaning of the words.

However, the preset program executed by the computing unit is pre-programmed in the computing unit and cannot be overwritten or changed, and the information that can be processed by the monitoring system is more diversified, and the preset program is relatively The complexity is higher, which increases the workload of the computing unit and reduces the overall computing performance of the monitoring system.

In view of this, the main purpose of this creation is to provide a parallel computing embedded sensing system to process multiple information in the shortest time and improve the overall computing performance of the monitoring system.

The embedded sensing system of the parallel computing of the present invention comprises: a sensing module for outputting a plurality of sensing signals; and a programmable computing module comprising: a front end signal processing unit including a analog/digital converter; Digitizing the sensing signals; and a programmable calculator comprising a plurality of computing units, each of the computing units being coupled to a memory, the programmable calculator receiving the sensing signals to perform a work row The work schedule includes a plurality of sequential execution phases, and the calculation results of the execution phase are stored in the memory as a data source for the next execution phase, wherein each execution phase includes at least one calculation unit for the The calculation unit calculates that the quantity of the at least one calculation unit is less than or equal to the number of the calculation units; and a transmission interface, and the calculation result of the programmable calculation module is transmitted externally.

According to the embedded sensing system of the parallel calculation of the present creation, the Reconfigurable calculator, as its name suggests, can schedule the work for the maximum performance of the system. In addition, the computing units of the programmable calculator are synchronously parallel. The calculation unit is calculated, so that each execution stage of the work schedule can be completed in the shortest time, and the requirement of processing the multi-information in the shortest time can be satisfied. Therefore, the transmission interface of the embedded sensing system of the present invention can be connected to a computer system to become a monitoring system, because a complicated operation flow has been completed in the programmable calculator of the present creation, and the computer system only receives the present. The calculation results of the creation, so the efficiency of the computer system will not be dragged down, greatly improving the overall computing performance of the monitoring system.

Referring to FIG. 1 , the embedded sensing system of the present parallel computing includes a sensing module 10 , a reconfigurable computing module 20 , and a transmission interface 30 .

The sensing module 10 is configured to output a plurality of sensing signals. For example, the sensing module 10 can include a plurality of sensors that sense ambient temperature, humidity, gas concentration, or sense respectively. Measuring the pulse signal of the human body, the electrocardiogram signal, or sensing the weight of the object, or sensing the pH value of the solution, etc., and the sensing signals output by the sensing module 10 can be voltage, current, resistance, and capacitance. Signals such as inductance, frequency, etc. are used to reflect the aforementioned ambient temperature, humidity, gas concentration, human body pulse signal, electrocardiogram signal, object weight, and pH value of the solution.

The input end of the programmable computing module 20 is connected to the output end of the sensing module 10 , and the programmable computing module 20 includes a front end signal processing unit 21 and a planability calculator 22 .

The front-end signal processing unit 21 includes an analog/digital converter 211, or may further include a filter 212 or a signal amplifier 213. The filter 212 and the signal amplifier 213 are sequentially connected to the analog/digital converter 211. After receiving the sensing signal of the sensing module 10, the device 212 and the signal amplifier 213 filter the noise of the sensing signal and amplify the signal strength, and the analog/digital converter 211 compares the senses of the analog type. The test signal is digitized and becomes a digital signal.

The planability calculator 22 includes a plurality of computing units 221, each of which is connected to a memory 222. After the planability calculator 22 receives the sensing signals from the front-end signal processing unit 21, according to the The sensing signal performs a work schedule, and the work schedule includes a plurality of sequential execution stages, and the data source of the current execution stage executed by the planability calculator 22 is the calculation result of the previous execution stage, or the The calculation result of the current execution phase performed by the planning calculator 22 serves as a data source for the next execution phase. Each of the execution stages includes one or more units of calculation, the number of units of calculation being less than or equal to the number of units of computing unit 221. When the execution stage of the planability calculator 22 includes a plurality of calculation units, the calculation units 221 simultaneously calculate the calculation units, and store the respective calculation results in the respective memory 222 as a lower a data source of an execution phase; and when the execution phase executed by the planability calculator 22 includes only one calculation unit, the calculation unit is calculated by one of the calculation units 221, and the calculation result is stored in Its memory serves as a source of information for the next implementation phase. It should be noted that the calculation units executed by the calculation unit 221 at each execution stage are pre-planned in the planability calculator 22.

The transmission interface 30 is connected to the planability calculator 22 to receive the calculation result of the planability calculator 22 and to externally transmit the calculation result. The transmission interface 30 can be a universal serial bus (USB) interface or a universal asynchronous transceiver module (UART) interface.

The work schedule performed by the planability calculator 22 will be described below by way of an example. Referring to FIG. 2, the planability calculator 22 includes a first calculating unit 223 and a second calculating unit 224. The first calculating unit 223 and the second calculating unit 224 are respectively connected to the memory 225 and 226. 3. The work schedule 40 includes a first execution phase A, a second execution phase B, and a third execution phase C. The first execution phase A includes a first calculation unit 41 and a second calculation unit 42. The second execution phase B includes a third calculation unit 43 and a fourth calculation unit 44, and the third execution phase C includes a fifth calculation unit 45.

In the first execution stage A, the first calculating unit 223 executes the first calculating unit 41 to receive two sensing signals a, b from the sensing module 10, and according to the two sensing signals a, b. A calculation result g is generated, and the calculation result g is stored in the memory 225 of the first calculation unit 223; the second calculation unit 224 executes the second calculation unit 42 to receive four strokes from the sensing module 10 Sensing signals c, d, e, f, and generating two calculation results h, i according to the four sensing signals c, d, e, f, and storing the two calculation results h, i in the second The memory 226 of the computing unit 224. It should be noted that, in the first execution stage A, the first calculating unit 223 and the second calculating unit 224 calculate the first calculating unit 41 and the second calculating unit 42 in parallel.

In the second execution phase B, the first calculating unit 223 executes the third calculating unit 43 to receive one of the sensing signals c and the calculation results g, h of the first execution phase A, and generate three calculation results s , k, z, and storing the calculation results s, k, z in the memory 225 of the first calculation unit 223; the second calculation unit 224 performs the fourth calculation unit 44 of the second execution phase B to receive The first execution stage A calculates the results h, i, and correspondingly generates two calculation results n, x, and stores the two calculation results n, x in the memory 226 of the second calculation unit 224. It should be noted that, in the second execution phase B, the first calculating unit 223 and the second calculating unit 224 calculate the third calculating unit 43 and the fourth calculating unit 44 in parallel.

In the third execution phase C, the first calculation unit 223 or the second calculation unit 224 executes the fifth calculation unit 45 of the third execution phase C to receive the calculation result z, x of the second execution phase B, and Correspondingly, a calculation result y is generated, and the calculation result y is stored in the memory 225, 226 of the first calculation unit 223 or the second calculation unit 224. The calculation result y of the third execution phase C is used as an output signal p, and the output signal p is transmitted to the transmission interface 30, and the output signal p is externally transmitted through the transmission interface 30.

According to the programmable calculator 22 of the present invention, the sensing module 10 generates a plurality of sensing signals, and the work schedule performed by the programmable calculator 22 is complicated because the programmable calculator is The computing units of 22 are synchronous parallel calculations, so that each execution phase of the work schedule can be completed in the shortest amount of time. Referring to FIG. 4, the embedded sensing system 50 of the present parallel computing can be connected to a local bus 51 of a computer system, the central processing unit 52 of the computer system, the main memory 53 and the input/output interface. The embedded sensing system 50 of the present parallel computing is connected through the local bus bar 51. Since the computer system is the calculation result of the embedded sensing system 50 that receives the parallel calculation of the present creation, in other words, the computer system does not need to be targeted to the sense. The plurality of sensing signals generated by the measuring module 10 are calculated, so that the burden of the computer system can be effectively reduced, and the computing performance of the computer system is relatively improved.

10‧‧‧Sensing module
20‧‧‧planable computing module
21‧‧‧ Front-end signal processing unit
211‧‧‧ Analog/Digital Converter
212‧‧‧ filter
213‧‧‧Signal Amplifier
22‧‧‧planable calculator
221‧‧‧Computation unit
222‧‧‧ memory
223‧‧‧First calculation unit
224‧‧‧Second calculation unit
225‧‧‧ memory
226‧‧‧ memory
30‧‧‧Transport interface
40‧‧‧Work schedule
41‧‧‧First unit of calculation
42‧‧‧Second calculation unit
43‧‧‧ Third calculation unit
44‧‧‧fourth calculation unit
45‧‧‧ fifth unit of calculation
50‧‧‧Parallel Computing Embedded Sensing System
51‧‧‧Local Busbars
52‧‧‧Central processor
53‧‧‧ main memory
54‧‧‧Input/Output Interface

Figure 1: Circuit block diagram of a preferred embodiment of the embedded sensing system of the present parallel computing. Figure 2: Circuit block diagram of the planability calculator in this creation. Figure 3: Flow chart of the work schedule performed by this creative. Figure 4: Schematic diagram of a parallel sensing embedded embedded sensing system connected to a computer system.

10‧‧‧Sensing module

20‧‧‧planable computing module

21‧‧‧ Front-end signal processing unit

211‧‧‧ Analog/Digital Converter

212‧‧‧ filter

213‧‧‧Signal Amplifier

22‧‧‧planable calculator

221‧‧‧Computation unit

222‧‧‧ memory

30‧‧‧Transport interface

Claims (4)

A parallel computing embedded sensing system includes: a sensing module for outputting a plurality of sensing signals; and a programmable computing module comprising: a front end signal processing unit including a analog/digital converter, Digitizing the sensing signals; and a programmable calculator comprising a plurality of computing units, each of the computing units being coupled to a memory, the programmable calculator receiving the sensing signals to perform a work row The work schedule includes a plurality of sequential execution phases, and the calculation results of the execution phase are stored in the memory as a data source for the next execution phase, wherein each execution phase includes at least one calculation unit for the The calculation unit calculates that the quantity of the at least one calculation unit is less than or equal to the number of the calculation units; and a transmission interface, and the calculation result of the programmable calculation module is transmitted externally. The parallel computing embedded sensing system of claim 1, wherein: the execution stage executed by the planability calculator includes a plurality of calculation units, and the calculation units respectively calculate the calculation units, and The respective calculation results are stored in the respective memory as the data source of the next execution stage; when the execution stage executed by the planability calculator includes only one calculation unit, the calculation unit calculates one of the calculation units Calculate the unit and store its calculations in its memory as a source of data for the next execution phase. The parallel-calculated embedded sensing system of claim 1 or 2, wherein the front-end signal processing unit further comprises a filter and a signal amplifier, wherein the filter, the signal amplifier and the analog/digital converter are sequentially connected. The parallel computing embedded sensing system according to claim 3, wherein the transmission interface is a universal serial bus interface or a universal asynchronous transceiver module interface.