Walking robot and control method thereof

A control method of a walking robot, the method comprising:

• performing transition of a first leg of the walking robot to a swing state and transition of a second leg of the walking robot to a support state;
• performing transition of the second leg from the support state to a toe-off state, when ground reaction force applied to the first leg exceeds a first set value and the first leg is in the swing state;
• performing transition of the second leg from the toe-off state to the swing state and transition of the first leg from the swing state to the support state, when ground reaction force applied to the second leg is below a second set value and the second leg is in the toe-off state;
• performing transition of the first leg from the support state to the toe-off state, when the ground reaction force applied to the second leg exceeds the first set value and the second leg is in the swing state;
• performing transition of the first leg from the toe-off state to the swing state, when the ground reaction force applied to the first leg is below the second set value and the first leg is in the toe-off state;
• and
• achieving walking of the walking robot by repeating the transitions of the first and second legs among the support state, the toe-off state, and the swing state,
• wherein torque τd(t) applied in the transition of the second leg of the walking robot to the toe-off state is defined by Expression 1,
• τd(t)=K′p(θd(t)−θ(t))−K′d{dot over (θ)}(t)+τff(t), Expression 1
• in Expression 1, K′p and K′d are gains, and τff(t) is feed forward torque.

A control method of a walking robot, the method comprising:

• performing transition of a second leg of the walking robot in a support state to a toe-off state, when ground reaction force applied to a first leg of the walking robot in a swing state exceeds a first set value;
• and
• performing transition of the second leg from the toe-off state to the swing state and transition of the first leg from the swing state to the support state, when ground reaction force applied to the second leg in the toe-off state is below a second set value,
• wherein torque τd(t) applied in the transition of the second leg of the walking robot to the toe-off state is defined by Expression 1,
• τd(t)=K′p(θd(t)−θ(t))−K′d{dot over (θ)}(t)+τff(t), Expression 1
• in Expression 1, K′p and K′d are gains, and τff(t) is feed forward torque.

A walking robot, comprising:

• a first leg;
• a second leg;
• a sensor unit configured to detect ground reaction forces respectively applied to the first leg and the second leg;
• and
• a control unit configured to perform transition of the second leg in a support state to a toe-off state, when the ground reaction force applied to the first leg in a swing state exceeds a first set value, and configured to perform transition of the second leg from the toe-off state to the swing state and transition of the first leg from the swing state to the support state, when the ground reaction force applied to the second leg in the toe-off state is below a second set value,
• wherein the control unit is configured to apply torque τd(t) defined by Expression 1 to the second leg in the toe-off state,
• τd(t)=K′p(θd(t)−θ(t))−K′d{dot over (θ)}(t)+τff(t), Expression 1
• in Expression 1, K′p and K′d are gains, and τff(t) is feed forward torque.