Code execution in real-time environments

[Gajendra1]

Code execution in real-time environments :
Real-Time Workshop offers you a comprehensive framework that allows you to execute code generated in real time and integrated into the execution environment. Depending on the model specified in the periodic sampling produces Real-Time Workshop rates either single-or multi-rate code. The program code is implemented with or without real-time operating system and runs either Singletasking, in multitasking or in asynchronous mode.


Singletasking :
In single-tasking mode, a simple scheduler calls the generated code on a single execution thread and thus prevents the task preemption between different sampling rates.

Multitasking :
In multitasking mode brings a deterministic scheduler ratenmontoner on the generated code, enabling the task preemption between different sampling rates. In bare-board environments, the preemption is done by nested interrupts. In RTOS environments to different priorities and provided the Task be used preemption.

Asynchronous execution :
In asynchronous mode is not periodic or asynchronous rates using Simulink S-functions are defined. Real-Time Workshop, these rates translated into the execution environment used tailored, target-specific code. Program code for trigger subsystems and for events such as hardware interrupts, in each case as a single, independent task model. For real-time operating system VxWorks own block library for asynchronous systems is provided, as you can use a template to create an environment suitable for your individual library.

[Gajendra1]

Real-time implementation :
Simulink and Real-Time Workshop target have a wide selection of independent controls for the real-time implementation. These include:

- Determining individual priorities for each sampling rate of a model
- Counter and timer in production quality, both the absolute and the elapsed time to calculate exactly
- A Rate Transition block to implement optimal data transfer mechanism between different sampling rates (such as semaphore, mutex and double buffering), which offer the best in each trade-off between data integrity, determinism and performance
- An overflow detection, to help create for each sample rate your own error handling logic can

Code Optimization :
Real-Time Workshop has several selectable optimizations that increase the efficiency of generated code. These include:
- Code reuse
- Folding of expressions
- Reuse of signal memory
- Elimination of dead paths
- Inlining Parameters
- Packaged and single-precision working target function libraries (for both ISO C and GNU C)

Working with Real-Time Workshop :
Real-Time Workshop is an integral part of the Simulink environment and is operated by the Simulink Model Explorer. All relevant parameters for code generation can be set to a central location. From the Model Explorer, you can:

- Code from Simulink models, or create individual sub
- Select Real-Time Workshop targets
- Targets set for code generation
- Different configuration records management

The built-in Simulink Model Advisor checks your model configuration and gives you on the basis of pre-defined goals or modeling guidelines and tips about optimizing the setup.

Definition and management of data :
With Real-Time Workshop, you have total control over how the data of a model in the generated code erscheinen.Zusätzlich it includes various options for data management. You can:

- Data types using declare built block data types (integer, floating point and fixed point)
- allocate space for the optimization and calibration of parameters
- allocate space for monitoring and recording of signal data
- Space reuse and minimize the local data storage

Real-Time Workshop to generate code from data that are stored either on the chart itself, but also in the Simulink Model Explorer provided data dictionary. So you can create various Data Dictionary for different targets and thus create a simple manner from the same model-specific code for many different platforms.