Custom Entity Generator Block with Signal Input and Signal Output

Step 4 of 6 in Create Custom Blocks Using MATLAB Discrete-Event System Block

This example shows how to create a custom source block that generates entities and to manage discrete states when implementing the discrete-event System object™ methods.

Suppose that you manage a facility that produces raw materials with a fixed deterministic rate. The materials contain a 12-digit bar code for stock management and priority values for order prioritization. To represent this behavior, this example shows how to generate a custom entity storage block is generated with one signal input port, one entity output port, one signal output port, and one storage element. The block generates entities with distinct priority values. The entities carry data and depart the block from its output port. The entity priority values are acquired from values of the incoming signal.

To open the model and to observe the behavior of the custom block, see CustomEntityGeneratorBlockExample.

Create the Discrete-Event System Object

Graphical representation of a Discrete-Event system showing a rectangle with an input port and two output ports containing a storage element. One of the cells of the storage elements points to its first cell through an arrow looping upwards labeled "Generate".

The block is defined as a custom entity generator block that generates entities with specified intergeneration periods. The generated entities carry data, and their priority values are determined by the values of the input signal.

See the Code to Create the Custom Entity Generator Block

classdef CustomEntityStorageBlockGeneration < matlab.DiscreteEventSystem...
                                            
    % A custom entity generator block.
    
    % Nontunable properties 
    properties (Nontunable)
        % Generation period
        period = 1;
    end

    properties(DiscreteState)
        % Entity priority
        priority;
        % Entity value
        value;
    end

    % Discrete-event algorithms
    methods
        function [events, out1] = setupEvents(obj)
            % Set up entity generation events at simulation start.
            events = obj.eventGenerate(1,'mygen',obj.period,obj.priority);
            % Set up the initial value of the output signal.
            out1 = 10;
        end

        function [entity,events,out1] = generate(obj,storage,entity,tag,in1)            
            % Specify event actions when entity is generated in storage.            
            entity.data = obj.value;
            % The priority value is assigned from the input signal.
            obj.priority = in1;
            % Output signal is the assigned priority value.
            out1 = obj.priority;
            events = [obj.eventForward('output',1,0) ...
                      obj.eventGenerate(1,'mygen',obj.period,obj.priority)];
        end
    end

    methods(Access = protected)
        
        function entityTypes = getEntityTypesImpl(obj)
            entityTypes = obj.entityType('Material');
        end
        
        function [inputTypes,outputTypes] = getEntityPortsImpl(obj)
            % Specify entity input and output ports. Return entity types at
            % a port as strings in a cell array. Use empty string to
            % indicate a data port.
            inputTypes = {''};
            outputTypes = {'Material',''};
        end

        function resetImpl(obj)
            % Initialize / reset discrete-state properties.
            obj.priority = 10;
            obj.value = 1:12;            
        end
        
        function [storageSpecs, I, O] = getEntityStorageImpl(obj)
            storageSpecs = obj.queueFIFO('Material', 1);
            I = 0;
            O = [1 0];
        end   

        function num = getNumInputsImpl(obj)
            % Define total number of inputs for system with optional
            % inputs.
            num = 1;
        end
        
        function num = getNumOutputsImpl(~)
            % Define total number of outputs.
            num = 2;
        end  
        function [out1 out2] = getOutputSizeImpl(obj)
            % Return size for each output port.
            out1 = [1 12];
            out2 = 1;
        end

        function [out1 out2] = getOutputDataTypeImpl(obj)
            % Return data type for each output port.
            out1 = "double";
            out2 = "double";
        end

        function [out1 out2] = isOutputComplexImpl(obj)
            % Return true for each output port with complex data.
            out1 = false;
            out2 = false;
        end

        function [sz,dt,cp] = getDiscreteStateSpecificationImpl(obj,name)
            % Return size, data type, and complexity of discrete-state
            % specified in name.
            switch name
                case 'priority'
                    sz = [1 1];
                case 'value'
                    sz = [1 12];
            end
            dt = "double";
            cp = false;           
        end
    end
end

Custom Block Behavior

Define the time between material generations.

    % Nontunable properties 
    properties (Nontunable)
        % Generation period
        period = 1;
    end

Initialize the discrete state variables.
```
    function resetImpl(obj)
        % Initialize / reset discrete-state properties.
        obj.priority = 10;
        obj.value = 1:12;            
    end
```
The variable priority represents material priority and the value represents bar code data carried by the materials.

Initialize the output for a source block.

    function num = getNumOutputsImpl(~)
        % Define total number of outputs.
        num = 2;
    end  
    function [out1 out2] = getOutputSizeImpl(obj)
    % Return size for each output port.
        out1 = [1 12];
        out2 = 1;
    end

    function [out1 out2] = getOutputDataTypeImpl(obj)
        % Return data type for each output port.
        out1 = "double";
        out2 = "double";
    end

    function [out1 out2] = isOutputComplexImpl(obj)
        % Return true for each output port with complex data.
        out1 = false;
        out2 = false;
    end

First function declares the output size.
Second function declares that output port data types are double.
Third function declares false for output ports because they do not support complex data.

Declare the size, data, and complexity of the discrete states.

    function [sz,dt,cp] = getDiscreteStateSpecificationImpl(obj,name)
        % Return size, data type, and complexity of discrete-state.
            switch name
                case 'priority'
                    sz = [1 1];
                case 'value'
                    sz = [1 12];
            end
            dt = "double";
            cp = false;           
    end

The discrete state priority is scalar. The data type is double and takes real values.
The discrete state value is a 1-by-12 vector. The data type is double and takes real values.

Generate the materials with intergeneration period, priority, and data defined by:

The parameter obj.period, declared as a public parameter that can be changed from the block dialog box.
The parameter obj.priority values, defined by the signal from the input port.
The parameter obj.value, a 1-by-12 vector which represents the data carried by entities.

    function events = setupEvents(obj)
        % Set up entity generation event for storage 1 at simulation start.
        events = obj.eventGenerate(1,'mygen',obj.period,obj.priority);
        % Set up the initial value of the output signal.
        out1 = 10;
    end

    function [entity,events,out1] = generate(obj,storage,entity,tag,in1)            
        % Specify event actions when entity is generated in storage.            
        entity.data = obj.value;
        % The value from the signal is assigned to the entity priority.
        obj.priority = in1;
        % Output signal is the assigned priority value.
        out1 = obj.priority;
        events = [obj.eventForward('output',1,0) ...
                    obj.eventGenerate(1,'mygen',obj.period,obj.priority)];
    end

Implement Custom Block

Save the .m file as CustomEntityStorageBlockGeneration. Link the System object to a SimEvents^® model by using a MATLAB Discrete-Event System block. For more information about linking, see Create Custom Blocks Using MATLAB Discrete-Event System Block.
Create a SimEvents model that includes the MATLAB Discrete-Event System block, a Ramp block, an Entity Terminator block, and two Scope blocks. Connect the blocks as shown in the model.
In the Ramp block, set Slope to 5 and Initial output to 10.
In the Entity Terminator block, you can display the priority values of the entities arriving at the block, in the Entry action field enter this code.
```
coder.extrinsic('fprintf');
fprintf('Priority: %d\n', double(entitySys.priority))
```
Right-click the entity path from the custom Entity Generator to the Entity Terminator and select the Log Selected Signals.
Simulate the model.
1. Observe the output of the Ramp block. For instance, the output value becomes 15, 20, 25, and 30 for the simulation time 1, 2, 3, and 4, respectively.
2. The Simulation Data Inspector shows that entities are forwarded to the Entity Terminator block with data of size 1-by-12.
3. You can also observe the priority values from the scope labeled Entity Priority for generation times 1, 2,3, 4, 5, 6, 7, 8, 9, and 10.