METHOD FOR CONFIGURING A TESTER EQUIPPED FOR TESTING A CONTROL UNIT

Abstract

A method for automated configuration of a tester equipped for testing a control unit. A first and second model of technical systems being executed in the tester. The execution of the models taking place periodically with defined sampling rates. An FPGA executes the first and/or the second model and a CPU executes the first or the second model. A first individual sampling rate is allocated for the first model and a second individual sampling rate is allocated for the second model. The first model is assigned for execution on either the CPU or the FPGA and the second model is assigned for execution on either the CPU or the FPGA. The tester is automatically configured for execution of the first model with the first allocated sampling rate on the FPGA or the CPU and of the second model with the second allocated sampling rate on the FPGA or the CPU.

Claims

1. A method for automated configuration of a tester equipped for testing a control unit, the method comprising: executing in the tester a first model of a technical system and a second model of a technical system, the execution of the models taking place periodically with defined sampling rates, the first model and the second model each representing a model of a technical system to be controlled or a model of a control unit under test, the first and second models being present as source code in a high-level programming language; providing in the tester an FPGA for executing the first and/or the second model and a CPU for executing the first or the second model; connecting the control unit under test to the tester during execution of the models and a data exchange takes place between the control unit and/or the first model and/or the second model; allocating a first individual sampling rate for the first model and allocating a second individual sampling rate for the second model; assigning the first model for execution on either the CPU or the FPGA and assigning the second model for execution on either the CPU or the FPGA; and configuring the tester for execution of the first model with the first allocated sampling rate on the FPGA or the CPU and of the second model with the second allocated sampling rate on the FPGA or the CPU.

2. The method according to claim 1, wherein the automatic configuration of the tester includes: generating at least one netlist from the first and/or the second model if execution on the FPGA was assigned to the first and/or the second model; generating communication interfaces to permit data exchange between the first model, the second model, and/or the control unit; configuring at least one partition of the FPGA by the at least one netlist and the generated communication interfaces; and configuring a multiple clock rate component on the FPGA to provide at least one individual clock rate signal for execution of the first and/or second model on the FPGA at the individual sampling rate.

3. The method according to claim 2, wherein the configuration of the tester includes: generating machine language source code from the first and/or the second model if execution on the CPU was assigned to the first and/or the second model; and linking the generated machine language source code to the configured FPGA via the generated communication interfaces.

4. The method according to claim 1, wherein the assignments of the models take place with the allocated sampling rates being taken into account.

5. The method according to claim 1, wherein the complexity of the models and/or free computing capacity on the CPU and/or the FPGA are taken into account during the assignment of the models.

6. The method according to claim 1, wherein the high-level programming language is a graphical programming language.

7. The method according to claim 2, wherein the generation of the netlist includes the intermediate step of generating a hardware description language.

8. The method according to claim 1, wherein the tester is a hardware-in-the-loop simulator or a rapid control prototyping system.

9. The method according to claim 1, wherein the tester has an input/output interface for connecting a control unit to the tester.

10. The method according to claim 2, wherein the configuration and the allocation of the sampling rates is done via an RCP/HIL hybrid configuration device, wherein the RCP/HIL hybrid configuration device is constituted such that it has a first serial arrangement of configuration elements for configuring an input/output interface and a second serial arrangement of configuration elements for configuring interfaces of the first and the second model, the configuration elements being assigned to one another, and the communication interfaces being generated while taking into account the configuration elements that are assigned to one another.

11. A tester for testing a control unit, the tester comprising: a first model of a technical system; a second model of a technical system, the first and second models being executed in the tester, the execution of the models being performed periodically with defined sampling rates, the first model and the second model each representing a model of a technical system to be controlled or a model of a control unit under test, the models being source code in a high-level programming language; an FPGA for executing the first and/or the second model; and a CPU for executing the first and/or the second model, wherein the tester is adapted to connect a control unit under test to the tester during execution of the models and adapted for data exchange between the control unit and/or the first model and/or the second model, and wherein the tester is adapted to execute the method according to claim 1.

12. A method for automated configuration of a tester equipped for testing a control unit, the method comprising: executing a first model of a technical system and a second model of a technical system in the tester, the execution of the first or second models being performed periodically with defined sampling rates, the first model and the second model each representing a model of a technical system to be controlled or a model of a control unit under test, the first and second models being source code in a high-level programming language; providing a FPGA in the tester for executing the first model and the second model; connecting the control unit under test to the tester during execution of the models such that a data exchange takes place between the control unit and/or the first model and/or the second model; allocating a first individual sampling rate for the first model; allocating a second individual sampling rate for the second model; configuring the tester for execution of the first model with the first allocated sampling rate and the second model with the second allocated sampling rate on the FPGA; generating at least one netlist from the first and the second model; generating communication interfaces to permit data exchange between the first model, the second model, and/or the control unit; configuring a first partition and a second partition of the FPGA by the at least one netlist and the generated communication interfaces; and configuring a multiple clock rate component on the FPGA to provide at least one first individual clock rate signal for execution of the first model with the first individual sampling rate and of a second individual clock rate signal for execution of the second model on the FPGA with the second individual sampling rate.

13. A method for automated configuration of a tester equipped for testing a control unit, the method comprising: executing a first model of a technical system and a second model of a technical system in the tester, the execution of the models being performed periodically with defined sampling rates, the first model and the second model each representing a model of a technical system to be controlled or a model of a control unit under test, the first model and the second model being source code in a high-level programming language; providing in the tester a first FPGA and a second FPGA for executing the first and/or the second model; connecting the control unit under test to the tester during execution of the models such that a data exchange takes place between the control unit and/or the first model and/or the second model; allocating a first individual sampling rate for the first model; allocating a second individual sampling rate for the second model; assigning the first model for execution on either the first FPGA or the second FPGA; assigning the second model for execution on either the first FPGA or the second FPGA with the allocated sampling rates being taken into account; and configuring the tester for execution of the first model with the first allocated sampling rate and of the second model with the second allocated sampling rate.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0028] The present invention will become more fully understood from the detailed description given hereinbelow and the accompanying drawings which are given by way of illustration only, and thus, are not limitive of the present invention, and wherein:

[0029] FIG. 1 shows an exemplary embodiment of a tester according to the invention;

[0030] FIG. 2 shows an FPGA that was configured using the method according to the invention for executing two models;

[0031] FIG. 3 depicts an exemplary form of an RCP/HIL hybrid configuration device 300; and

[0032] FIG. 4 shows an exemplary graphical programming language.

DETAILED DESCRIPTION

[0033] Shown in FIG. 1 is an exemplary embodiment of a tester according to the invention 100, to which a control unit 200 is connected through an input/output interface 102. The input/output interface can have several different hardware channels, for instance digital or analog inputs and outputs at which electrical signals can be received or transmitted. A first model (103) of a technical system and a second model (104) of a technical system can be executed on the tester. In this context, a technical system can be, e.g., an engine of a motor vehicle, a motor vehicle itself or parts thereof, a control unit, or any other desired technical system. In order to execute the models, the tester contains various execution resources, such as, e.g., an FPGA 105 and/or a CPU 106. The execution resources need not necessarily be present directly in the tester, but instead can also be externally connected to the tester, for example in the form of embedded systems or plug-in cards. In the embodiment shown here, the execution resources, which is to say here a CPU 106 and an FPGA 105, are connected to a communication interface 107, for example a data bus, for data exchange with one another. This communication interface can additionally be connected to the input/output interface 103 as well, in order to thus establish communication of the models with the control unit connected to the tester.

[0034] The tester 100 can be, for example, a hardware in the loop (HIL) simulator. The tester 100 can also be a rapid control prototyping (RCP) system. However, the tester can also be a device that is suitable for performing HIL tests or RCP tests due to the fact that a model of a technical system can be executed on the tester, and this model can exchange data through input/output interfaces with a device under test that is connected to the tester, e.g., a control unit, with this data exchange being used to analyze, in particular, the reaction of the tester to data resulting from the model, which data are transmitted to the control unit in the form of electrical signals, for example.

[0035] A technical model, which is to say a model of a technical system, can by way of example. be present in the form of a software model that is specified by source code, for example in a high-level language such as C, C++, or in a machine language such as, e.g., assembler or executable machine code. Using a technical model, any desired systems can be modeled in order to simulate them virtually. Thus, for example, a model of a motor can be present as software, with the software being programmed such that during a simulation, which is to say here an execution of the model on a CPU or an FPGA, input parameters are processed by the software and output values are generated as a function of the input parameters and the nature of the model. An input parameter in this context can be, for instance, a voltage present at a throttle valve of a gasoline engine, and output values in this regard could be a resultant opening angle of the throttle valve, fuel consumption, and/or a resultant torque at the crankshaft. The model can also be a model of a control unit under test or under development, however.

[0036] FIG. 1 shows one possible configuration of a tester 100 after a method according to the invention in accordance with claim 1 has been carried out. To execute a model on the tester, an individual sampling rate is allocated to the model. Instead of a sampling rate, a sampling time can of course also be allocated in accordance with the invention, wherein the sampling rate represents the inverse of the sampling time. This allocation of the sampling rate takes place for each model that is executed on the tester. The allocation can take place manually or automatically. In the example from FIG. 1, a sampling time of 10 ms can be allocated to the first model 103 and a sampling time of 20 ms can be allocated to the second model 104, for example. Accordingly, the first model calculates new output values of the model every 10 ms, whereas the second model only calculates new output values every 20 ms with which electrical signals can then be generated that can be sent to the control unit 200. Depending on the complexity of the model, it is not always possible for an execution resource of the tester to calculate the model in the allocated sampling time. In such a case, an assignment of the model must take place for execution on a resource of the tester that has the appropriate computing performance to calculate the model in the predetermined sampling time. Thus, in the example the execution resource FPGA can be assigned to the first model 103 and the execution resource CPU can be assigned to the second model. According to the invention, an automatic configuration of the tester additionally takes place so that the tester is enabled to execute the models on the tester in such a manner that they can communicate with a device under test connected to the tester within the framework of a simulation; in the specific case of the example, this is thus a configuration of the tester for executing the first model 103 on the FPGA 105 with the sampling time 10 ms and a configuration of the tester for execution of the second model 104 on the CPU 106 with the sampling time 20 ms.

[0037] According to the invention, the allocation of the models to an execution resource can also be carried out automatically taking into account the allocated sampling rates, the complexity of the calculations performed in the model, and the data types used, as well as the already existing utilization of the execution resources or computing capacities, e.g. in the form of task turnaround times of a CPU or free logic gates of an FPGA. For this purpose, a multivariable optimization algorithm taking into account secondary conditions and a quadratic objective function can be used, for example. The objective function here can contain a maximal latency, which is to say a reaction time of the models to a change in the input values. In addition, the objective function can be extended by criteria that take into account the allocated sampling rates and the complexity of the calculations as well as the data types used. The secondary conditions in the optimization algorithm can ensure that task turnaround times of the CPU are adhered to and the FPGA contains sufficient free logic gates.

[0038] The automatic configuration can also include an automatic generation of netlists for the particular models that are to be executed on an FPGA. Various software solutions with automation interfaces, such as, e.g., the Xilinx System Generator, Matlab HDL Coder, and Xilinx Vivado HLS, can be used for this purpose, for example. From models that are present in a high-level programming language such as graphical Simulink code or C/C++, for example, these software solutions can automatically generate netlists or hardware description languages such as VHDL or Verilog code with which FPGAs can be configured to execute the program specified in the high-level programming language.

[0039] The communication interface 107 can also be generated automatically in accordance with the invention. In contrast to the form shown in FIG. 1, it can also be constituted such that it enables the communication between two models that are executed on one and the same FPGA. The communication interface 107 can also be constituted as a combined interface, which means that the communication interface is composed of multiple independent interfaces. The communication interface or even parts of a combined communication interface can be represented in a graphical programming environment as model port input and output blocks by means of which the automatic generation of the communication interfaces can be configured with regard to a data exchange.

[0040] FIG. 2 shows an FPGA 105, which, in contrast to the example shown in FIG. 1, was configured automatically such that both the first model 103 and the second model 104 can be executed on the FPGA 105. For parallel execution of the models on the same FPGA, one partition is configured for execution of the first model and a second partition is configured for execution of the second model. So that each of the two models 103 and 104 can be executed with individual clock rates on the FPGA, the models receive a separate clock rate signal from a multiple clock rate component. The multiple clock rate component, in turn, receives a clock rate signal from the original clock rate component 202 permanently implemented on the FPGA, which signal is transformed by the multiple clock rate component into the individual clock rate signals 203, for example into one clock rate signal that permits a sampling time of 10 ms and one clock rate signal that permits a sampling time of 20 ms. It must be noted here that the multiple clock rate component 201 is not limited to the generation of two clock rate signals, but rather is fundamentally capable of generating any arbitrary number of clock rate signals. The two models 103 and 104 can exchange data by means of the communication interface 107 during their execution, both with one another and with other execution resources of additional models being executed and with a device 200 under test connected to the tester. The multiple clock rate component can also include mechanisms for enabling debugging functions. With these mechanisms it is possible, for example, to generate a clock signal that can be used for debugging functionalities.

[0041] The allocation of the sampling rates can be accomplished according to the invention by means of an RCP/HIL hybrid configuration device 300, for example, which is diagrammed in FIG. 3. An RCP/HIL hybrid configuration environment is distinguished by the fact that RCP and/or HIL simulators can be configured with it. The RCP/HIL hybrid configuration device in FIG. 3 contains a first serial arrangement of configuration elements 301 and 302 for configuring the input/output interface of the tester 100. In addition, the RCP/HIL hybrid configuration device 300 contains a second serial arrangement of configuration elements 303 and 304 for configuring interfaces of the first and the second model. To configure the communication interface, the configuration elements 301 and 302 for configuration of the input/output interface of the tester 100 can be assigned to the configuration elements 303 and 304 for configuration of interfaces of the first and of the second model, which can be accomplished through a connection with connecting elements 305, for example. Moreover, to configure the communication interface, the configuration elements 303 and 304 for configuration of interfaces of the first and of the second model can be mutually assigned to one another, for example through a connection with an additional connecting element 306, by which means a data exchange between the models 103 and 104 can be specified.

[0042] FIG. 4 shows an exemplary form of a graphical programming language. Here, program components and their functional relationships are configured through graphical elements, such as blocks and lines, for example. The graphical program 400 in FIG. 4 can be constituted such that, for example, said graphical program reads in a first input value 401 and a second input value 402, then further processes the read data in accordance with the functionality specified in the block 403 in order to output, by means of the output block 404, an output value calculated by means of the graphical program. The functionality of the block 403 could include, for instance, in adding or multiplying the input values, in differentiating or integrating them, or the like. The graphical blocks can be connected to one another by connecting lines in order to thus establish a data exchange between blocks. Of course, a graphical program 400 is not limited to the embodiment shown here, but instead can be significantly more complex in nature, which can be reflected in the number of blocks and connections of the blocks. The graphical program 400 can be used for modeling a technical system. Consequently, a graphical program 400 can also be a model of a technical system 103 and/or 104.

[0043] The invention being thus described, it will be obvious that the same may be varied in many ways. Such variations are not to be regarded as a departure from the spirit and scope of the invention, and all such modifications as would be obvious to one skilled in the art are to be included within the scope of the following claims.